2025年9月28日 (日)

第517回:フェアユースに関する2023年5月18日のアンディ・ウォーホル事件アメリカ最高裁判決

 前回、AI学習へのフェアユースの適用について述べる3つのアメリカ地裁判決を紹介したが、これらの地裁判決やアメリカ著作権局のAI学習と著作権の関係に関する報告書(第512回参照)にも大きな影響を与えているものとして、最も最近のフェアユース最高裁判例である2023年5月18日のアンディ・ウォーホル財団対事件アメリカ最高裁判決を避けて通る事はできないので、遅ればせではあるが、ここでもこの判決について少し取り上げておきたいと思う。(なお、これより1つ前の、JavaのAPI利用に関する2021年4月5日のグーグル対オラクル事件アメリカ最高裁判決については第440回参照。)

 この事件は、ポップアートで有名なアンディ・ウォーホル氏の作品である「オレンジ・プリンス」(歌手のプリンス氏の写真をオレンジ色に着色したシルクスクリーン作品)の利用がフェアユースである事を確認するために氏の作品を管理しているアンディ・ウォーホル財団が提起したものであって、ニューヨーク州南部地区地裁はフェアユースの適用を認めたが、その後、第2巡回控訴裁がそれを覆し、最高裁まで持ち込まれ、結果として控訴裁の判断が支持され、最終的にフェアユースは認めなかったというものである(事件については、英語版Wikipediaも参照)。

 まず、この最高裁判決(pdf)から、要旨の結論部分を以下に訳出する。(いつも通り、翻訳は全て拙訳である。)

Held: The "purpose and character" of AWF's use of Goldsmith's photograph in commercially licensing Orange Prince to Conde Nast does not favor AWF's fair use defense to copyright infringement. Pp. 12-38.

...

(c) Goldsmith's original works, like those of other photographers, are entitled to copyright protection, even against famous artists. Such protection includes the right to prepare derivative works that transform the original. The use of a copyrighted work may nevertheless be fair if, among other things, the use has a purpose and character that is sufficiently distinct from the original. In this case, however, Goldsmith's photograph of Prince, and AWF's copying use of the photograph in an image licensed to a special edition magazine devoted to Prince, share substantially the same commercial purpose. AWF has offered no other persuasive justification for its unauthorized use of the photograph. While the Court has cautioned that the four statutory fairuse factors may not "be treated in isolation, one from another," but instead all must be "weighed together, in light of the purposes of copyright," Campbell, 510 U. S., at 578, here AWF challenges only the Court of Appeals' determinations on the first fair use factor, and the Court agrees the first factor favors Goldsmith. P. 38.

判示:オレンジ・プリンスのコンデ・ナストへの商業的な許諾という、ゴールドスミスの写真のAWF(アンディ・ウォーホル財団)による利用の「目的及び性質」は、著作権侵害に対するフェアユースの抗弁においてAWFに有利なものではない。

(略)

(c)他の写真家のものと同様、ゴールドスミスの元の作品は、有名な芸術家に対しても、著作権により保護され得る。この保護は元のものを変形する派生作品を用意する権利を含む。とりわけ利用が元のものから十分に区別される目的及び性質を有している場合、著作物の利用はそれでも公正なものであり得る。しかしながら、この事件においては、ゴールドスミスによるプリンスの写真と、プリンスに捧げられた雑誌の特別号に許諾された画像における写真のAWFによる複製利用は実質的に同じ商業的目的を共有する。写真の無許可利用に対し、AWFは他に納得の行く正当化事由を持ち出せていない。裁判所は、4つの法定のフェアユースの要素は「1つ1つ孤立して扱われ」得ず、「著作権の目的に照らし、一緒に衡量され」なくてはならないものであること、1994年3月7日のキャンベル対エイカフ=ローズ・ミュージック社事件アメリカ最高裁判決、を注意して来たが、ここで、AWFは控訴裁の第1のフェアユースの要素に関する判断のみに異議を申し立てている所、当裁判所も第1の要素はゴールドスミスに有利である事に同意する。

 この事件は利用における変形性が十分でなかったためにフェアユースが認められなかったものとして取り上げられる事もなくはないのだが、この要旨の部分でも書かれている様に、元々ウォーホル氏がオレンジ・プリンスに用いたゴールドスミス氏の写真について特定の雑誌への掲載用に1回限りの許諾・ライセンスを受けており、その後、ウォーホル財団が同作品の他の雑誌への利用をゴールドスミス氏の許諾なしで行ったという、かなり特殊な事情が背景にある事に注意が必要なものと私は考えている。

 さらに判決本文のソトマイヨール判事他の多数意見から、この事情と変形利用について、特にポイントとなる事が書かれていると私が考える部分を抜き出すと以下の様になる。

...

Taken together, these two elements - that Goldsmith's photograph and AWF's 2016 licensing of Orange Prince share substantially the same purpose, and that AWF's use of Goldsmith's photo was of a commercial nature - counsel against fair use, absent some other justification for copying. That is, although a use's transformativeness may outweigh its commercial character, here, both elements point in the same direction.

The foregoing does not mean, however, that derivative works borrowing heavily from an original cannot be fair uses. In Google, the Court suggested that "[a]n 'artistic painting' might, for example, fall within the scope of fair use even though it precisely replicates a copyrighted 'advertising logo to make a comment about consumerism.'" 593 U. S., at ___-___ (slip op., at 24-25) (quoting 4 M. Nimmer & D. Nimmer, Copyright §13.05[A][1][b] (2019), in turn quoting N. Netanel, Making Sense of Fair Use, 15 Lewis & Clark L. Rev. 715, 746 (2011) (some internal quotation marks omitted)). That suggestion refers to Warhol's works that incorporate advertising logos, such as the Campbell's Soup Cans series. See fig. 7, infra.

Yet not all of Warhol's works, nor all uses of them, give rise to the same fair use analysis. In fact, Soup Cans well illustrates the distinction drawn here. The purpose of Campbell's logo is to advertise soup. Warhol's canvases do not share that purpose. Rather, the Soup Cans series uses Campbell's copyrighted work for an artistic commentary on consumerism, a purpose that is orthogonal to advertising soup. The use therefore does not supersede the objects of the advertising logo.

...

Campbell did describe a transformative use as one that "alter[s] the first [work] with new expression, meaning, or message." 510 U. S., at 579; see also Google, 593 U. S., at ___ (slip op., at 24). That description paraphrased Judge Leval's law review article, which referred to "new information, new aesthetics, new insights and understandings." Leval 1111. (Judge Leval contrasted such additions with secondary uses that "merely repackag[e]" the original. Ibid.) But Campbell cannot be read to mean that §107(1) weighs in favor of any use that adds some new expression, meaning, or message. Otherwise, "transformative use" would swallow the copyright owner's exclusive right to prepare derivative works. Many derivative works, including musical arrangements, film and stage adaptions, sequels, spinoffs, and others that "recast, transfor[m] or adap[t]" the original, §101, add new expression, meaning or message, or provide new information, new aesthetics, new insights and understandings. That is an intractable problem for AWF's interpretation of transformative use. The first fair use factor would not weigh in favor of a commercial remix of Prince's "Purple Rain" just because the remix added new expression or had a different aesthetic. A film or musical adaptation, like that of Alice Walker's The Color Purple, might win awards for its "significant creative contribution"; alter the meaning of a classic novel; and add "important new expression," such as images, performances, original music, and lyrics. Post, at 11, 23 (KAGAN, J., dissenting) (internal quotation marks omitted). But that does not in itself dispense with the need for licensing.

...
(略)

合わせて考えると、これらの2つの事項-ゴールドスミスの写真とAWFの2016年のオレンジ・プリンスの許諾行為が実質的に同じ目的を共有している事と、AWFによるゴールドスミスの写真の利用が商業的性質のものである事-はフェアユースに反する方向に働くものであり、複製に対する何かしらの他の正当化事由をを欠く。つまり、利用の変形性がその商業的性格を上回り得るにせよ、ここでは、2つの事項は同じ方向を向いている。

上記の事は、しかしながら、元の作品から多くを借りる派生作品はフェアユースではあり得ないという事を意味しない。2021年4月5日のグーグル対オラクル事件アメリカ最高裁判決において、当裁判所は、「例えば、著作権の保護を受ける『広告ロゴを消費主義に対する批評をするために』正確に複写したとしても、芸術的な絵画はフェアユースに該当し得る」事を示唆した。グーグル事件判決(ニンマー著作権法を引用、次にN.ネタネル著、フェアユースの理解を引用)。この示唆は、キャンベルのスープ缶シリーズの様な、広告ロゴを取り入れたウォーホルの作品を指すものである。

ただし、全てのウォーホルの作品が、そのあらゆる利用が同じフェアユース分析をもたらすものではない。実際の所、スープ缶はここで引かれる区別をはっきりと示すものである。キャンベルのロゴの目的はスープを広告する事である。ウォーホルのキャンバスはこの目的を共有しない。むしろ、スープ缶シリーズはキャンベルの著作物を、スープの広告と直交する目的である、消費主義に対する芸術的批評のために用いている。したがって、この利用は広告ロゴの対象を奪うものではない。

(略)

キャンベル事件アメリカ最高裁判決は、変形利用を「最初のもの[作品]を新しい表現、意味又はメッセージとともに変化させた」ものと記載した。グーグル事件アメリカ最高裁判決も参照。この記載は、「新しい情報、新しい美学、新しい知見及び理解」に言及した、レヴァル判事の法学論文、フェアユースの基準に向けて、を言い換えたものである。(レヴァル判事は、この様な追加と元のものを「単に再パッケージしただけ」の2次利用を対比している。)しかし、キャンベルはアメリカ著作権法第107条第1項が、新しい表現、意味又はメッセージを追加するあらゆる利用に有利に衡量される事を意味すると読んではならない。さもなくば、「変形利用」は派生作品を用意する著作権者の排他的権利を飲み込んでしまうであろう。音楽のアレンジ、映画化や劇場化、続編、スピンオフ及び元のものを「作り直すか、変形するか又は適応させる」その他のものを含む、多くの派生作品は、新しい表現、意味又はメッセージを追加するか、新しい情報、新しい美学、新しい知見及び理解を提供するものである。これが変形利用のAWFの解釈に関わる厄介な問題である。そのリミックスが新しい表現を追加するか異なる美学を有するだけで、第1のフェアユースの要素がプリンスの「パープル・レイン」の商業的なミックスに有利に衡量される事はないであろう。映画化や劇場化は、アリス・ウォーカーのザ・カラー・パープルの様に、その「大きな創作的寄与」のために受賞し得るであろう;古典小説の意味を変化させ得るでろう;そして、画像、実演、オリジナルの音楽及び歌詞の様な、「重要な新しい表現」を追加し得るであろう。後の意見(ケイガン判事が反対)。しかし、この事はそれ自体で許諾の必要性を免除しない。

(略)

 また、この多数意見に賛成しながら、ゴーサッチ判事他が補足意見として、以下の様に述べている事も、この判決を理解する上では重要だろう。

Last but hardly least, while our interpretation of the first fair-use factor does not favor the Foundation in this case, it may in others. If, for example, the Foundation had sought to display Mr. Warhol's image of Prince in a nonprofit museum or a for-profit book commenting on 20th-century art, the purpose and character of that use might well point to fair use. But those cases are not this case. Before us, Ms. Goldsmith challenges only the Foundation's effort to use its portrait as a commercial substitute for her own protected photograph in sales to magazines looking for images of Prince to accompany articles about the musician. And our only point today is that, while the Foundation may often have a fair-use defense for Mr. Warhol's work, that does not mean it always will. Under the law Congress has given us, each challenged use must be assessed on its own terms.

最後に重要な事であるが、この事件において、第1のフェアユースの要素の私たちの解釈は財団に有利とならないものの、それは他の場合には有利となり得る。例えば、もし財団がウォーホル氏のプリンスの画像を非営利の美術館又は営利の20世紀美術の批評本で示そうとしたのなら、利用の目的及び性質はフェアユースの方に大きく向かうであろう。しかし、この事件はこの様な場合ではない。私たちの前で、ゴールドスミス氏は、歌手に関する記事に合わせプリンスの画像を求める雑誌に販売するためにその保護を受ける写真の商業的代替物としてその肖像を利用しようとする財団の活動に対抗しているに過ぎない。そして、私たちの今日の要点は、財団はウォーホル氏の作品についてフェアユースの抗弁を多く持ち出せるだろうが、これは常にそうできる事を意味しないという事である。アメリカ著作権法の下でアメリカ議会が私たちに与えたのは、訴えられたそれぞれの利用がそれぞれ自身の条件において評価されなければならないという事である。

 この判決はアメリカの最高裁の判決であり、その多数意見がアメリカの著作権判例として確立されたと見なければならないのだが、そうだとしても、この事件の背景として元の著作者であるウォーホル氏が写真家のゴールドスミス氏からその写真の利用について雑誌に対するライセンスを受けており、後にウォーホル財団が全く同様な雑誌への利用のライセンスを行ったという事情が重きをなしている事を抜きにする事はできないだろう。第1の要素においてフェアユースを否定する方向の結論を導くために一面としてその様な事が書かれているのも確かだが、ライセンスに関する事情を抜きにこの判例を一般化して、フェアユースと認められるためにはウォーホル氏の作品の変形性が足りなかったという事のみを言おうとするのは危険なミスリードであって、この判決の射程は狭く捉えるべきと私は考えている。

 さらに書いておくと、私自身は、この事件において、本当にライセンスに関する事情がフェアユースの判断に影響を与えるべきか、ウォーホル氏の作品の変形性が不十分かという点について甚だ疑問に思っており、かえって以下のケイガン判事他の反対意見の方が正しいと思っている。

...

Now recall all the ways Warhol, in making a Prince portrait from the Goldsmith photo, "add[ed] something new, with a further purpose or different character" - all the way she "alter[ed] the [original work's] expression, meaning, [and] message." Ibid. The differences in form and appearance, relating to "composition, presentation, color palette, and media." 1 App. 227; see supra, at 7-10. The differences in meaning that arose from replacing a realistic - and indeed humanistic - depiction of the performer with an unnatural, disembodied, masklike one. See ibid. The conveyance of new messages about celebrity culture and its personal and societal impacts. See ibid. The presence of, in a word, "transformation" - the kind of creative building that copyright exists to encourage. Warhol's use, to be sure, had a commercial aspect. Like most artists, Warhol did not want to hide his works in a garret; he wanted to sell them. But as Campbell and Google both demonstrate (and as further discussed below), that fact is nothing near the showstopper the majority claims. Remember, the more transformative the work, the less commercialism matters. See Campbell, 510 U. S., at 579; supra, at 14; ante, at 18 (acknowledging the point, even while refusing to give it any meaning). The dazzling creativity evident in the Prince portrait might not get Warhol all the way home in the fair-use inquiry; there remain other factors to be considered and possibly weighed against the first one. See supra, at 2, 10, 14. But the "purpose and character of [Warhol's] use" of the copyrighted work - what he did to the Goldsmith photo, in service of what objects - counts powerfully in his favor. He started with an old photo, but he created a new new thing.

II

The majority does not see it. And I mean that literally. There is precious little evidence in today's opinion that the majority has actually looked at these images, much less that it has engaged with expert views of their aesthetics and meaning. Whatever new expression Warhol added, the majority says, was not transformative. See ante, at 25. Apparently, Warhol made only "modest alterations." Ante, at 33. Anyone, the majority suggests, could have "crop[ped],flatten[ed], trace[d], and color[ed] the photo" as Warhol did. Ante, at 8. True, Warhol portrayed Prince "somewhat differently." Ante, at 33. But the "degree of difference" is too small: It consists merely in applying Warhol's "characteristic style" - an aesthetic gloss, if you will - "to bring out a particular meaning" that was already "available in [Goldsmith's] photograph." Ibid. So too, Warhol's commentary on celebrity culture matters not at all; the majority is not willing to concede that it even exists. See ante, at 34 ("even if such commentary is perceptible"). And as for the District Court's view that Warhol transformed Prince from a "vulnerable, uncomfortable person to an iconic, larger-than-life figure," the majority is downright dismissive. Ante, at 32. Vulnerable, iconic - who cares? The silkscreen and the photo, the majority claims, still have the same "essential nature." Ante, at 25, n. 14 (emphasis deleted).

The description is disheartening. It's as though Warhol is an Instagram filter, and a simple one at that (e.g., sepia-tinting). "What is all the fuss about?," the majority wants to know. Ignoring reams of expert evidence - explaining, a severy art historian could explain, exactly what the fuss is about - the majority plants itself firmly in the "I could paint that" school of art criticism. No wonder the majority sees the two images as essentially fungible products in the magazine market - publish this one, publish that one, what does it matter? See ante, at 22-23; supra, at 10. The problem is that it does matter, for all the reasons given in the record and discussed above. See supra, at 9-10. Warhol based his silkscreen on a photo, but fundamentally changed its character and meaning. In belittling those creative contributions, the majority guarantees that it will reach the wrong result.

Worse still, the majority maintains that those contributions, even if significant, just would not matter. All of Warhol's artistry and social commentary is negated by one thing: Warhol licensed his portrait to a magazine, and Goldsmith sometimes licensed her photos to magazines too. That is the sum and substance of the majority opinion. Over and over, the majority incants that "[b]oth [works] are portraits of Prince used in magazines to illustrate stories about Prince"; they therefore both "share substantially the same purpose" - meaning, a commercial one. Ante, at 22-23, 38; see ante, at 12-13, 27, n. 15, 33, 35. Or said otherwise, because Warhol entered into a licensing transaction with Conde Nast, he could not get any help from factor 1 - regardless how transformative his image was. See, e.g., ante, at 35 (Warhol's licensing "outweigh[s]" any "new meaning or message" he could have offered). The majority'scommercialism-trumps-creativity analysis has only one way out. If Warhol had used Goldsmith's photo to comment on or critique Goldsmith's photo, he might have availed himself of that factor's benefit (though why anyone would be interested in that work is mysterious). See ante, at 34. But because he instead commented on society - the dehumanizing culture of celebrity - he is (go figure) out of luck.

From top-to-bottom, the analysis fails. It does not fit the copyright statute. It is not faithful to our precedent. And it does not serve the purpose both Congress and the Court have understood to lie at the core of fair use: "stimulat[ing]creativity," by enabling artists and writers of every description to build on prior works. Google, 593 U. S., at ___ (slip op., at 24). That is how art, literature, and music happen; it is also how all forms of knowledge advance. Even as the majority misconstrues the law, it misunderstands - and threatens - the creative process.

...

III

And the workings of creativity bring us back to Andy Warhol. For Warhol, as this Court noted in Google, is the very embodiment of transformative copying. He is proof of concept - that an artist working from a model can create important new expression. Or said more strongly, that appropriations can help bring great art into being. Warhol is a towering figure in modern art not despite but because of his use of source materials. His work - whether Soup Cans and Brillo Boxes or Marilyn and Prince - turned something not his into something all his own. Except that it also became all of ours, because his work today occupies a significant place not only in our museums but in our wider artistic culture. And if the majority somehow cannot see it - well, that's what evidentiary records are for. The one in this case contained undisputed testimony, and lots of it, that Warhol's Prince series conveyed a fundamentally different idea, in a fundamentally different artistic style, than the photo he started from. That is not the end of the fair-use inquiry. The test, recall, has four parts, with one focusing squarely on Goldsmith's interests. But factor 1 is supposed to measure what Warhol has done. Did his "new work" "add[]something new, with a further purpose or different character"? Campbell, 510 U. S., at 579. Did it "alter[] the first with new expression, meaning, or message"? Ibid. It did, and it did. In failing to give Warhol credit for that transformation, the majority distorts ultimate resolution of the fair-use question.

Still more troubling are the consequences of today's ruling for other artists. If Warhol does not get credit for trans-formative copying, who will? And when artists less famous than Warhol cannot benefit from fair use, it will matter even more. Goldsmith would probably have granted Warhol a license with few conditions, and for a price well within his budget. But as our precedents show, licensors sometimes place stringent limits on follow-on uses, especially to prevent kinds of expression they disapprove. And licensors may charge fees that prevent many or most artists from gaining access to original works. Of course, that is all well and good if an artist wants merely to copy the original and market it as his own. Preventing those uses - and thus incentivizing the creation of original works - is what copyrights are for. But when the artist wants to make a trans-formative use, a different issue is presented. By now, the reason why should be obvious. "Inhibit[ing] subsequent writers" and artists from "improv[ing] upon prior works" - as the majority does today - will "frustrate the very ends sought to be attained" by copyright law. Harper & Row, 471 U. S., at 549. It will stifle creativity of every sort. It will impede new art and music and literature. It will thwart the expression of new ideas and the attainment of new knowledge. It will make our world poorer.

(略)

さて、しっかり思い出そう、ウォーホルは、ゴールドスミスの写真からプリンスの肖像を作るにあたり、「さらなる目的又は異なる性質をもって何がしか新しいものを追加した」のであり、それは完全に「[元の作品の]表現、意味、[及び]メッセージを変えた」のである。キャンベル事件アメリカ最高裁判決。「構成、表示、色彩及び媒体」に関し、形態と外観の違いがあり。付録;上記。現実的な-そして実際人間的な-実演家の描写を不自然で、被実態的で、仮面の様なものと置き換える事から生じる意味の違いがある。同上。セレブ文化とその個人的、社会的影響に対する新しいメッセージの伝達がある。同上。要するに、「変容」-これを促進するために著作権が存在している類の創作的構築-があるのである。ウォーホルの利用は、確かに、商業的側面を有する。ほとんどの芸術家同様、ウォーホルはその作品を屋根裏部屋に隠しておく事を望まず;売る事を望んだ。しかし、キャンベル事件アメリカ最高裁判決及びグーグル事件アメリカ最高裁判決がともに示している通り、この事実は多数意見が主張する様な拍手喝采の決定事項ではない。作品がより変形的であれば、商業的である事はより問題とならない事を思い起こすべきである。キャンベル事件アメリカ最高裁判決参照;上記;前の意見(それに意味を与える事を拒みながらも、この点を認めている)。プリンスの肖像において明らかな素晴らしい創造性はフェアユースの調査においてウォーホルを完全に安全なものとしないであろう;考慮されるべき他の要素が残り、これらは第1の要素に反する様に衡量され得る。上記。しかし、-彼がゴールドスミスの写真に対してなした-著作物の「[ウォーホルの]利用の目的と性質」は強力に彼に有利なものと考えられる。彼は古い写真から出発したが、新しく新しい物を作り出したのである。

多数意見はこの事を見ていない。そして、私は文字通りそうであると言う。今日の意見で多数意見が実際にこれらの画像を詳細に見たという証拠はたかが知れており、その美学と意味において専門家の意見を取り入れたという事はさらにない。ウォーホルが追加した新しい表現は何であれ変形的ではないと多数意見は言う。ウォーホルは「ささやかな修正」をしただけである。前の意見。誰でもウォーホルがした様に「写真をトリミングし、平坦化し、なぞり、色づけする」事ができたと多数意見は示唆する。前の意見。真実、ウォーホルはプリンスを「何がしか別の形で」肖像化した。前の意見。しかし、「違いの程度」は小さ過ぎ:-そうしようとするのは、美学的こじつけであろうが-それはウォーホルの「典型的なスタイル」を適用し、「[ゴールドスミスの]写真に」既にある「特定の意味を浮き彫りにしようとした」に過ぎない。同上。そうしてまた、ウォーホルのセレブ文化に対する批評は何ら問題とはならない;多数意見はそれが存在する事すら譲ろうとはしていない。前の意見参照(「たとえその様な批評が感得し得るとしても」)。そして、ウォーホルが「脆弱で、落ち着かない人間を偶像的な人生以上の像」に変形したという地裁の見解についても多数意見は徹底的に拒否している。前の意見。脆弱、偶像的-誰が気にするのか?シルクスクリーンと写真はなお同じ「本質的な性質」を持っていると多数意見は主張する。前の意見。

この記載は心温まるものではない。それはウォーホルはインスタグラムのフィルターであり、その様に単純なものである(例えば、セピア着色)と言う様なものである。「それが一体何だと言うのか?」、多数意見は訊ねる。大量の専門家の証拠-厳格な芸術の歴史家が説明できるだろう様に、正確にそれが何なのかを説明している-を無視し、多数意見は芸術表の「私もそれを描けた」派に強固に根ざしている。多数意見が2つの画像を雑誌の市場において本質的に代替可能なものと見ているのは不思議ではない-一方を掲載し、他を掲載する、それになんの問題があるだろうか?前の意見;上記参照。記録にあり、上で議論した全ての理由から、それが問題になるというのが実際問題なのである。ウォーホルはそのシルクスクリーンで写真を基礎としたが、本質的にその性質と意味を変えた。この創作的寄与を軽んじ、多数意見が誤った結果に至るのは約束されている。

さらに悪い事に、多数意見は、この寄与が、大きなものであったとしても、問題とならないだろう事を主張している。あらゆるウォーホルの芸術家的、社会的批評は1つの事によって否定される:ウォーホルはその肖像を雑誌にライセンスし、ゴールドスミスもしばしばその写真を雑誌にライセンスしたという事。これが多数意見の要旨であり本質である。繰り返し、多数意見は「両方[の作品]はともにプリンスに関する物語を示すために雑誌において使われるプリンスの肖像である」と唱える;したがって、それらはともに「本質的に目的」-目的として商業的なもの-を共有すると。前の意見参照。別の言い方をするなら、ウォーホルがコンデ・ナストとライセンス契約を結んだため、-その画像がどれほど変形的であるかに関わらず-彼は要素1から何の助けも得られないという事である。例えば、前の意見参照(ウォーホルのライセンス行為は如何なる「新しい意味又はメッセージ」をも上回る)。多数意見の消費主義が切り札となる創造性の分析は1つの出口しかない。もしウォーホルがゴールドスミスの写真の批評又は批判のためにゴールドスミスの写真を利用したなら、この要素の利益を得られたであろう(何故誰もがこの作品に興味を引かれるのは謎であるが)。前の意見参照。しかし、彼は代わりに社会-セレブの脱人間化文化-を批判したために彼は(奇妙な事に)運を逃した。

最初から最後まで、その分析は間違っている。それは著作権法の成文に合致しない。私たちの先例に忠実でもない。そして、アメリカの議会と裁判所がともにフェアユースの根幹にあると理解して来た、先の作品に基づきあらゆる表現を構築する事を芸術家と作家に可能にする事により「創造性を喚起する」という目的に資するものでもない。グーグル事件アメリカ最高裁判決。これが芸術、文学及び音楽の発生の仕方であり;あらゆる形式の知識の進展の仕方である。多数意見は法の解釈を誤る事で、それは創作プロセスを誤解し-さらに脅かしているのである。

(略)

そして、創造性の働きから私たちはアンディ・ウォーホルに戻る。グーグル事件アメリカ最高裁判決で当裁判所が記した様に、ウォーホルは変形複製の正しく化身だからである。彼は-芸術家の仕事がモデルから重要な新しい表現を作り出し得る事の-概念実証である。さらに強く言うなら、この流用は偉大な芸術をもたらす助けともなり得る。ウォーホルはその元となるマテリアルの利用に関わらずではなくそのために現代美術において聳え立つ人物となっている。彼の作品は-スープ缶とブリロの箱、マリリンとプリンスのいずれでも-彼のものではない何かをすっかり彼自身の何かに変えるものである。その事を除いても、それは全て私たちのものとなっている、彼の作品は今日私たちの美術館だけでなく私たちのより広い文化において大きな地位を占めているのであるから。そして、多数意見がこの事を分からないとしても-そう、証拠の記録はそう言っている。この事件のものは争いのない証言を含んでおり、その多くは、ウォーホルのプリンスシリーズは彼が出発した所の写真とは本質的に異なる観念を本質的に異なる芸術的スタイルで伝えるものであると言っている。思い起こそう、テストは4つの部分からなり、1つでゴールドスミスの利益に集中している。しかし、要素1はウォーホルがした事を測る事が想定されるものである。彼の「新しい作品」は「さらなる目的又は異なる性質を持つ、何か新しいものを追加した」か?キャンベル事件アメリカ最高裁判決。それは「最初のものを新しい表現、意味又はメッセージで変化させた」か?同上。その通り、それはそうしたのである。その変形においてウォーホルに信用を与えない事で、多数意見はフェアユースの課題の究極的な解決を捻じ曲げている。

さらに困る事は、今日の判決の他の芸術家に対する影響である。もしウォーホルがその変形複製に対して信用を得られないとしたら、誰ができるだろうか?そして、ウォーホルほど有名でない芸術家がフェアユースから利益を得られない時、それはさらに問題となるだろう。ゴールドスミスは多少の条件とともに十分彼の予算の範囲内の価格でウォーホルにライセンスを与えたのであろう。しかし、私たちの先例が示す通り、ライセンサーが特に同意しない種類の表現を抑止するために後の利用について厳しい制限を課す事は良くある。そして、元の作品へのアクセスを得る事を多くの又はほとんどの芸術家に抑止するためにライセンサーが料金を課す事もある。もちろん、芸術家が元のものを複製し、自身の市場で販売する事だけを求めるのであれば、これは全く良い事である。この様な利用を抑止する事-そして、その様にして元の作品の創作にインセンティブを与える事-は著作権の目的である。しかし、芸術家が変形利用をする事を求める時、異なる事柄が現れる。今、その理由は明らかに違いない。-今日多数意見がそうした様に-「後に続く作家」及び芸術家が「先の作品に基づいて改善するのを禁じる事は」著作権法が達しようとする「正にその目的を妨げる」だろう。1985年5月20日のハーパー&ロー出版社対ナショナル・エンタープライズ事件アメリカ最高裁判決。それはあらゆる種類の創造性を窒息させるだろう。それは新しい芸術、音楽及び文学の妨げとなるだろう。それは新しいアイデアの表現と新しい知識の獲得を阻害するだろう。それは私たちの世界をより貧しいものとするだろう。

 この判決の射程は狭く捉えるべきとしても、反対意見の中の、アンディ・ウォーホル氏の変形利用においてフェアユースが認められなければ一体誰に認められるのかという問いこそ非常に本質的な問いである。省略した部分で反対意見は丁寧に美術の歴史を辿っているが、美術の文脈の下で考えれば、「オレンジ・プリンス」も含め、元の写真や作品の表現がなお残っているにせよ、アンディ・ウォーホル氏はそこから自分自身の作品を作り出しているのであって、この様な作品こそ最も典型的にフェアユースが認められて然るべきものに違いないのである。(なお、今回フェアユースと日本の著作権法第30条の4の非享受目的利用のための権利制限との関係について細かく述べる事はしないが、元の著作物の表現が多少享受可能な形で残るにせよ、パロディも含め、実質的に異なるものと言えるレベルに達している、この様な変形利用作品こそ、私が日本でも権利制限の対象に追加するべきと常々考えているものである。)

 そこで過去のライセンスに関する事情が何か影響するだろうかと考えると、これも反対意見の様に、甚だ疑問である。ウォーホル氏が最初にライセンスした訳ではないが適法に入手した他のソースから写真を抜き出して加工したとしたらどうだろうか。それでフェアユースどうかが変わって良いだろうか。さらに、ウォーホル氏の作品の独自性から考えて、雑誌の市場においてウォーホル氏の作品が使えなければ代わりにゴールドスミス氏の作品を使うだろうという様な代替性があるかも怪しい。

 もちろん程度問題であって、多少の変形があったとしても、実質フリーライドといって良いような複製行為についてフェアユースを認めるべきでないという事もその通りであるが、このアンディ・ウォーホル事件に関する限り、反対意見の方が妥当であると、将来のフェアユース判例において見直しが図られるべきと私は考えている。

 繰り返しになるが、多数意見が判例として通用すると言っても、アメリカ最高裁の多数意見がこのアンディ・ウォーホル事件でフェアユースを認めないという結論に至ったのは、ウォーホル氏が最初に利用した写真について雑誌向けのライセンス契約をしており、後にウォーホル財団が氏の作品の利用を同様に雑誌にライセンスしたという事情があってこそのものに違いない。さらに、争われたのが第1の要素のみであり、その中でほぼ全てを検討しなければならなかったという事もあるだろう。この様な事情を抜きにして作品の変形性の程度・十分性についてのみ判例を一般化して扱う事は危険なミスリードであって、この判決の射程は狭く捉えるべきものと私は考えている。

 すなわち、このアンディ・ウォーホル事件アメリカ最高裁判決は、最近の人工知能(AI)の学習への著作物の利用との関係で取り上げられる事もあり、下級審の判例やアメリカ著作権局の報告書にも変形性に重きを置く様な形で影響を与えている様に見えるのだが、この様に、本来芸術の文脈の下に考えられるべきだったにも関わらず、ライセンスに関する事情の方ばかりを見てフェアユースに関する判断を下した判決からその事情を抜きにして変形性について一般化を行い無理に適用しようとするべきではないというのが私の考えである。さらに言うと、AI学習への著作物の利用においても何かしら既存のライセンスはあり得るだろうが、フェアユース分析はその事情のみに引き擦られるべきでもない。AI学習とフェアユースの関係を問う事件がアメリカ最高裁まで上がった時には、4つのフェアユースの要素が全て評価され、その中でAI学習との関係でも新しい作品を作り出すための著作物の利用は原則フェアユースであると認められる事を私は期待している。

| | コメント (0)

2025年8月24日 (日)

第516回:人工知能(AI)の学習への著作物の利用のフェアユース該当性に関する3つのアメリカの地裁判決

 ここ半年くらいの間に、まだ地裁レベルだが、人工知能(AI)の学習への著作物の利用のフェアユース該当性に関する判決がアメリカで3つ出されている。

 アメリカのフェアユースは、

  1. 利用の目的及び性質
  2. 著作物の性質
  3. 利用された著作物の部分の量と本質性
  4. 著作物の価値と市場への影響

という4つの要素の比較衡量に基づき、その利用が公正かどうかを判断する著作権の一般例外条項であり、特にこれらの要素の判断を述べている部分を中心に、今回はそれぞれの判決の内容を見て行く。

(1)2025年2月11日のトムソン・ロイター社対ロス・インテリジェンス社事件デラウェア地裁判決

 最初の事件は、最近良く取沙汰される生成AIが問題になったものではないが、トムソン・ロイターが作った判決の要旨を学習データに使って利用者に質問に対して類似する判決などを教えるが、元の要旨そのものを出力するものではない、リサーチツールの提供がフェアユースに該当するかが争われたものである。

 デラウェア地裁は、2025年2月11日の判決(pdf)で、以下の様な判断により、この様な利用はフェアユースに該当しないとした。(以下、翻訳は全て拙訳。)

A. Factor one goes to Thomson Reuters

First, I consider the purpose and character of Ross's use. 17 U.S.C. § 107(1). I look mainly at whether it was commercial and whether it was transformative. Andy Warhol Found. for the Visual Arts, Inc. v. Goldsmith, 598 U.S. 508, 529-31 (2023). If Ross and Thomson Reuters use copyrighted material like the headnotes for very similar purposes and Ross's use is commercial, this factor likely disfavors fair use. Id. at 532-33.

1. Ross's use is commercial. Ross admits as much. D.I. 727 at 29. It "stands to profit from exploitation of the copyrighted material without paying the customary price." Harper & Row, 471 U.S. at 562. But commerciality is not dispositive. I must balance it against how different this work's purpose or character is. Warhol, 598 U.S. at 525.

2. Ross's use is not transformative. Transformativeness is about the purpose of the use. "If an original work and a secondary use share the same or highly similar purposes, and the second use is of a commercial nature, the first factor is likely to weigh against fair use, absent some other justification for copying." Warhol, 598 U.S. at 532-33. It weighs against fair use here. Ross's use is not transformative because it does not have a "further purpose or different character" from Thomson Reuters's. Id. at 529.

Ross was using Thomson Reuters's headnotes as AI data to create a legal research tool to compete with Westlaw. It is undisputed that Ross's AI is not generative AI (AI that writes new content itself). Rather, when a user enters a legal question, Ross spits back relevant judicial opinions that have already been written. D.I. 723 at 5. That process resembles how Westlaw uses headnotes and key numbers to return a list of cases with fitting headnotes. Thomson Reuters uses its headnotes and Key Number System primarily to help legal researchers navigate Westlaw and (possibly, as the parties dispute this) to improve Westlaw's internal search tool. D.I. 769 at 14(10:24:52). The parties agree that Ross and Westlaw are competitors. D.I. 752-1 at 4. So at first glance, this factor looks simple.

...

My prior opinion wrongly concluded that I had to send this factor to a jury. 694 F. Supp. 3d at 483-84. I based that conclusion on Sony and Sega. Since then, I have realized that the intermediate-copying cases (1) are computer-programming copying cases; and (2) depend in part on the need to copy to reach the underlying ideas. Neither is true here. Because of that, this case fits more neatly into the newer framework advanced by Warhol. I thus look to the broad purpose and character of Ross's use. Ross took the headnotes to make it easier to develop a competing legal research tool. So Ross's use is not transformative. Because the AI landscape is changing rapidly, I note for readers that only non-generative AI is before me today.

...

B. Factor two goes to Ross

Second, I ask about the nature of the original work. 17 U.S.C. § 107(2). That involves "focus[ing] on the degree of creativity inherent to the work." 4 Nimmer on Copyright § 13F.06. More creative works get more protection. Id. § 13F.06[A].

Westlaw's material has more than the minimal spark of originality required for copyright validity. But the material is not that creative. Though the headnotes required editorial creativity and judgment, that creativity is less than that of a novelist or artist drafting a work from scratch. And the Key Number System is a factual compilation, so its creativity is limited.

I signaled a similar leaning before. 694 F. Supp. 3d at 484-85. Yet I stopped short of granting summary judgment based on factual disputes about how much creativity was involved. Now, as I concluded above, there is no factual dispute that the headnotes have creative elements but are far from the most creative works.

So factor two goes for Ross. Note, though, that this factor "has rarely played a significant role in the determination of a fair use dispute." Authors Guild, 804 F.3d at 220.

C. Factor three goes to Ross

Third, I focus on how much of the work was used and how substantial a part it was relative to the whole. 17 U.S.C. § 107(3). I ask whether that usage was "reasonable in relation to the purpose of the copying." Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 586 (1994). Courts consider both "the quantity of the materials used" and "their quality and importance." Id. at 587. To win on this factor, the alleged copier must not take the "heart" of the work. Id.

My prior opinion did not decide factor three but suggested that it leaned towards Ross. The opinion focused on Ross's claim that its output to an end user is a judicial opinion, not a West headnote, so it "communicates little sense of the original." 649 F. Supp. 3d at 485 (quoting Authors Guild, 804 F.3d at 223).

I stand by that reasoning, but now go a step further and decide factor three for Ross. There is no factual dispute: Ross's output to an end user does not include a West headnote. What matters is not "the amount and substantiality of the portion used in making a copy, but rather the amount and substantiality of what is thereby made accessible to a public for which it may serve as a competing substitute." Authors Guild, 804 F.3d at 222 (internal quotation marks omitted). Because Ross did not make West headnotes available to the public, Ross benefits from factor three.

...

D. Factor four goes to Thomson Reuters

Factor four "is undoubtedly the single most important element of fair use." Harper & Row, 471 U.S. at 566. For this factor, I consider the "likely effect [of Ross's copying] on the market for the original." Campbell, 510 U.S. at 590. I must consider not only current markets but also potential derivative ones "that creators of original works would in general develop or license others to develop." Id. at 592. I also consider any "public benefits the copying will likely produce." Google, 593 U.S. at 35. The original market is obvious: legal-research platforms. And at least one potential derivative market is also obvious: data to train legal AIs.

My prior opinion left this factor for the jury. I thought that "Ross's use might be transformative, creating a brand-new research platform that serves a different purpose than Westlaw." 694 F. Supp. 3d at 486. If that were true, then Ross would not be a market substitute for Westlaw. Plus, I worried whether there was a relevant, genuine issue of material fact about whether Thomson Reuters would use its data to train AI tools or sell its headnotes as training data. Id. And I thought a jury ought to sort out "whether the public's interest is better served by protecting a creator or a copier." Id.

In hindsight, those concerns are unpersuasive. Even taking all facts in favor of Ross, it meant to compete with Westlaw by developing a market substitute. D.I. 752-1 at 4. And it does not matter whether Thomson Reuters has used the data to train its own legal search tools; the effect on a potential market for AI training data is enough. Ross bears the burden of proof. It has not put forward enough facts to show that these markets do not exist and would not be affected.

Nor does a possible benefit to the public save Ross. Yes, there is a public interest in accessing the law. But legal opinions are freely available, and "the public's interest in the subject matter" alone is not enough. Harper & Row, 471 U.S. at 569. The public has no right to Thomson Reuters's parsing of the law. Copyrights encourage people to develop things that help society, like good legal-research tools. Their builders earn the right to be paid accordingly. This case is distinguishable from Google, where the API was valuable "because users, including programmers, [were] just used to it." 593 U.S. at 38. There is nothing that Thomson Reuters created that Ross could not have created for itself or hired LegalEase to create for it without infringing Thomson Reuters's copyrights.

E. Balancing the factors, I reject Ross's fair-use defense

Factors one and four favor Thomson Reuters. Factors two and three favor Ross. Factor two matters less than the others, and factor four matters more. Weighing them all together, I grant summary judgment for Thomson Reuters on fair use.

A.要素1はトムソン・ロイターに有利

第1に、私はロスの利用の目的と性質を考慮する。アメリカ著作権法第107条第1項。主としてそれが商業的なものか及びそれが変形的なものであるかを見て行く。2023年5月18日のアンディ・ウォーホル財団対ゴールドスミス事件アメリカ最高裁判決。ロス及びトムソン・ロイターが要旨の様な著作権の保護を受けるマテリアルを非常によく似た目的に利用し、ロスの利用が商業的なものである場合、この要素はフェアユースに不利なものとなり得るであろう。同上。

1.ロスの利用は商業的である。ロスはその通り認めている。事件記録。それは「通常の料金を支払う事なく著作権の保護を受けるマテリアルの利用から利益を受ける立場にある。」1985年5月20日のハーパー&ロー出版社対ナショナル・エンタープライズ事件アメリカ最高裁判決。しかし、商業性は決定的なものではない。それとこの著作物の目的及び性質がどれほど異なるのかとのバランスが考えられなければならない。ウォーホル事件アメリカ最高裁判決。

2.ロスの利用は変形的なものではない。変形性は利用の目的に関するものである。「元の著作物及び2次利用が同じまたは高度に類似した目的をともにし、2次利用が商業的な性質のものである場合、コピーに対する他の正当化事由がない限り、第1の要素はファユースに反する方向に働くであろう。」ウォーホル事件アメリカ最高裁判決。ここでそれはフェアユースに反する方向に働く。トムソン・ロイターの利用から見てさらなる利用又は異なる性質を持つものではないため、ロスの利用は変形的なものではない。同上。

ロスはウェストローと競合する法的リサーチツールを作成するためにAIデータとしてトムソン・ロイターの要旨を利用していた。ロスのAIが生成AI(それ自体で新しいコンテンツを書くAI)でない事には争いがない。むしろ、利用者が法的質問を入力する時、ロスは既に書かれた関係する法的意見を吐き出す。事件記録。そのプロセスは、ウェストローが合致する要旨とともに事件のリストを返すために要旨及びキーナンバーを使うやり方と類似する。トムソン・ロイターはその要旨及びキーナンバーシステムを主として法律研究者がウェストローを見て回るため、そして、(この事については当事者が争っているので、恐らく)ウェストローの内部サーチツールを改善するために使っている。事件記録。当事者はロスとウェストローが競争関係にある事を認めている。事件記録。したがって、一見、この要素は単純に見える。

(略)

私の前の意見でこの要素を陪審に送るべきと結論づけたのは間違っていた。2023年9月25日の意見。私のその結論は2000年2月10日のソニー・コンピュータ・エンタテイメント社対コネティックス社事件第9巡回控訴裁判決及び1992年10月20日のセガ・エンタープライゼス社対アコレード社事件第9巡回控訴裁判決(訳注:いずれも互換性のためのソフトウェアのリバースエンジニアリングはフェアユースに該当し得るとした事件)に基づいていた。その時、私は、(1)中間的コピーが問題となる事件はコンピュータプログラミングのコピーが関わるものであり;(2)基礎的なアイデアに到達するためにコピーする必要性に一部依存すると考えていた。ここではいずれも正しくない。そのため、本事件にはウォーホル事件によって進められたより新しい枠組みの方がより適切に合致する。ここで、私はロスの利用の広い目的と性質に着目する。ロスは競合する法的リサーチツールの開発をより容易にするために要旨を取った。よって、ロスの利用は変形的なものでない。AIを巡る状況は急速に変化しており、今私の前にあるのは生成的なものでないAIである事を読者に注意しておく。

(略)

B.要素2はロスに有利

第1に、私は元の著作物の性質について検討する。アメリカ著作権法第107条第2項。これは「著作物の本質的な創作性の程度への注目」を含む。ニンマー著作権法。より創作的な著作物がより保護を受けられる。同上。

ウェストローのマテリアルは有効な著作権に求められる独創性の最低限の発露以上のものを有する。しかし、このマテリアルはそれほど創作的なものではない。要旨には編集的な創作性と判断が求められるが、この創作性は一から著作物を作り上げる小説家や芸術家のものより低い。そして、キーナンバーシステムは事実の編集であり、その創作性は限定的なものである。

私は似た傾向を前に示していた。2023年9月25日の意見。ただし、どれほどの創作性が含まれているかという事実に関する争いに基づき略式判決を出す事はしなかった。今は、上で結論づけた通り、要旨は創作的要素を持つが最も創作的な著作物からは程遠い事という事実に関する争いはない。

よって、要素2はロスに有利である。ただし、この要素は「フェアユースに関する争いの決定において重大な役割を果たす事は稀である」事を注記しておく。2015年10月16日の全米作家協会対グーグル社事件第2巡回控訴裁判決。

C.要素3はロスに有利

第3に、私はは著作物のどれほどが使われたか及その部分が全体に対してどれほど本質的かに注目する。アメリカ著作権法第107条第3項。私はどれほどその利用が「コピーの目的との関係で合理的」かを検討する。1994年3月7日のキャンベル対エイカフ=ローズ・ミュージック社事件アメリカ最高裁判決。裁判所は「利用されたマテリアルの量」及び「その質と重要性」をともに考慮する。同上。この要素において勝つためには、被疑複製者は著作物の「核心」を取ってはならない。

私の前の意見は要素3について決定しなかったが、ロスに向かう傾向を示唆していた。その意見は、エンドユーザーに対するその出力は法的意見であって、ウェストの要旨ではなく、よって、それは「元の意味をほとんど伝えていない」というロスの主張に重点を置いていた。2023年9月25日の意見(全米作家協会事件第2巡回控訴裁判決を引用)。

私はこの理由を支持するが、さらに一歩を進め、要素3はロス有利と決定する。次の事実に対する争いはない:エンドユーザーに対するロスの出力はウェストの要旨を含まない。問題となるのは「コピーを作成する際に利用された部分の量及び本質性」ではなく、むしろ「競合する代替物として提供される先の公衆にそれによってアクセス可能なものとされるものの量と本質性」である。全米作家協会事件第2巡回控訴裁判決。ロスはウェストの要旨を公衆に入手可能としていなかったのであるから、ロスは要素3において利がある。

(略)

D.要素4はトムソン・ロイターに有利

要素4は「疑いの余地なくフェアユースの唯一最も有用な要素である。」ハーパー&ロー事件アメリカ最高裁判決。この要素について、私は「[ロスのコピーの]元のものに関する市場へのあり得る影響」を考慮する。キャンベル事件アメリカ最高裁判決。私は現在の市場のみならず、「元の作品の創作者が通常展開するか他の者に展開させる事を許諾する」潜在的派生市場も考慮しなければならない。同上。私はまた「複製行為が作り出すであろう公共の利益」も考慮する。2021年4月5日のグーグル社対米オラクル社事件アメリカ最高裁判決。元の市場は明らかである:法的リサーチプラットフォーム。そして、少なくとも1つの潜在的派生市場もまた明らかである:法的AIの学習のためのデータ。

私の前の意見はこの要素を陪審の判断に委ねようとするものだった。私は「ロスの利用は変形的であり、ウェストローとは異なる目的に用いられる全く新しいリサーチプラットフォームを作り出すものであり得る」と考えていた。2023年9月25日の意見。それが本当なら、ロスはウェストローの代替市場とならないであろう。さらに、トムソン・ロイターがそのデータをAIツールの学習に用いたかその要旨を学習データとして売ったであろうかという具体的な事実に関する本質的な問題があるかどうかという事を懸念していた。同上。そして、「創作者か複製者かを守る事によって公共の利益にどちらがより資するか」を陪審が整理すべきと考えていた。同上。

後になって思えば、これらの懸念は納得しかねるものであった。ロスに有利な様に全ての事実を取ったとしても、それは代替市場を展開する事からウェストローと競合する事になる筈である。事件記録。そして、トムソン・ロイターが自身の法的サーチツールの学習にそのデータを用いたかどうかは問題とならない;AI学習データのための潜在的市場に対する影響で十分である。ロスが立証責任を負う。これらの市場が存在しておらず、また影響も受けないだろう事を示す十分な事実を提示していない。

あり得る公衆への利益もロスを救う事はない。確かに、法へのアクセスには公共の利益がある。法的意見は自由に入手可能であるが、「実質的な事項における公共の利益」はそれだけで十分なものではない。ハーパー&ロー事件アメリカ最高裁判決。公衆はトムソン・ロイターの法の解析に対して何ら権利を有さない。著作権は良い法的リサーチツールの様な社会を助けるものを開発する事を人々に促すものである。その構築者はそれに応じて支払いを受ける権利を得る。本事件は、「プログラマーを含む利用者が丁度それを使う事から」APIが価値のあるものであったグーグル事件とは区別されるものである。トムソン・ロイターが作り出したもので、トムソン・ロイターの著作権を侵害する事なく、ロスがそれ自体で作り出し得なかったか、自身のためにリーガルイーズを雇って作り出したかしたものは何もない。

E.要素を比較衡量し、私はロスのフェアユースの抗弁を斥ける

要素1及び4はトムソン・ロイターに有利である。要素2及び3はロスに有利である。要素2は他のものより重要でなく、要素4はより重要である。これらを全て衡量し、私はフェアユースについてトムソン・ロイターに利があるとの略式判決を与える。

 上で翻訳した部分を読めば大体分かると思うが、要するに、この事件では、創作性の程度は低いにしても著作権の保護を受ける判決の要旨が、全く同種の競合リサーチツールを作るために使われていたので、その利用が変形的と考えられない事が第1の要素において、リサーチツール等の市場に影響を与え得る事が第4の要素において重視され、フェアユース該当性が否定されている。

(2)2025年6月23日のアンドレア・バーツ他対アンスロピック社事件カリフォルニア北部地裁判決

 次のものは大規模言語モデル(LLM)を用いた所謂生成AIが問題となったアンドレア・バーツ他対アンスロピック社事件におけるで、2025年6月23日のカリフォルニア北部地裁の判決である。

 裁判所は、その判決(pdf)で、以下の様に、新しい文章を作り出す生成AIのための著作物の利用は基本的に変形的なものであってフェアユースたり得るが、元の著作物の入手が合法か違法かによって判断は変わるとした。

1. THE PURPOSE AND CHARACTER OF THE USE.

...

A. THE COPIES USED TO TRAIN SPECIFIC LLMS.

...

A better analogue to our facts would be an AI tool trained - using court opinions, and briefs, law review articles, and the like - to receive legal prompts and respond with fresh legal writing. And, on facts much like those, a different court came out the other way. It found fair use. White v. W. Pub. Corp., 29 F. Supp. 3d 396, 400 (S.D.N.Y. 2014) (Judge Jed Rakoff).

The latter use stood sufficiently "orthogonal" to anything that any copyright owner rightly could expect to control. See Warhol, 598 U.S. at 538-40. It could thus be freed up for the copyist to use, "promot[ing] the progress of science and the arts, without diminishing the incentive to create." Id. at 531 (emphasis added); see U.S. CONST. art. I, § 8, cl. 8.

In short, the purpose and character of using copyrighted works to train LLMs to generate new text was quintessentially transformative. Like any reader aspiring to be a writer, Anthropic's LLMs trained upon works not to race ahead and replicate or supplant them - but to turn a hard corner and create something different. If this training process reasonably required making copies within the LLM or otherwise, those copies were engaged in a transformative use.

The first factor favors fair use for the training copies.

B. THE COPIES USED TO BUILD A CENTRAL LIBRARY.

But that is not the only use at issue. Recall that Anthropic purchased millions of print books for its central library and pirated millions of digital books for its central library, too. It used specific sets and subsets of books for training specific LLMs. And, it then retained all the copies in its central library for other uses that might arise even after deciding it would not use them to train any LLM (at all or ever again). Anthropic seems to believe that because some of the works it copied were sometimes used in training LLMs, Anthropic was entitled to take for free all the works in the world and keep them forever with no further accounting. There is no carveout, however, from the Copyright Act for AI companies.

Because the legal issues differ between the library copies Anthropic purchased and pirated, this order takes them in turn.

(i) The Purchased Library Copies Converted from Print to Digital.

...

In sum, the first fair use factor favors fair use for the digital library copies converted from purchased print library copies - but these do not excuse the pirated library copies.

(ii) The Pirated Library Copies.

...

In sum, the first factor points against fair use for the central library copies made from pirated sources - and no damages from pirating copies could be undone by later paying for copies of the same works.

2. THE NATURE OF THE COPYRIGHTED WORK.

...

Here, Anthropic accepts that all of Authors' books - all published, whether non-fiction or fiction - contained expressive elements (Reply 9). And, as set out above, this order accepts Authors' view of the evidence that their works were chosen for their expressive qualities in building a central library and then in training specific LLMs (Opp. 11, 17 (citing, e.g., Opp. Exh. 3 at -03433)).

...

The second factor points against fair use for all copies alike.

3. THE AMOUNT AND SUBSTANTIALITY OF THE PORTION USED.

...

A. THE COPIES USED TO TRAIN SPECIFIC LLMS.

...

"What matters [ ] is not so much 'the amount and substantiality of the portion used' in making a copy, but rather the amount and substantiality of what is thereby made accessible to a public [in the purported secondary use] for which it may serve as a competing substitute [for the primary use]." Google, 804 F.3d at 222. Here, once again, there is no allegation of any traceable connection between the Claude service's outputs and Authors' works. The copying used to train the LLMs underlying Claude was thus especially reasonable.

...

For one thing, all agree Anthropic needed billions of words to train any given LLM. If using only books, Anthropic would have needed millions of books per model. If using a set comprising only a small fraction of books and a larger fraction of other texts, Anthropic still would have needed hundreds of thousands of books. Authors contend that because Anthropic showed it could use such smaller sets of books, it surely could have used no books at all - or at least not their books (Opp. 23). But Authors forget that "reasonably necessary" does not mean "strictly necessary." Authors do not contest that the volume of text required to train an LLM is monumental. Because using so many works was reasonably necessary, using any one work for actually training LLMs was about as reasonable as the next.

For another thing, no output to the public was even alleged to be infringing. So, yes, Authors' works were chosen as the strongest examples of writing. But the compelling benefits of training the LLMs on strong examples were not offset by revelations to the public of any portion of the works themselves. What was copied was therefore especially reasonable and compelling.

The third factor thus favors fair use for the training copies.

B. THE COPIES USED TO BUILD A CENTRAL LIBRARY.

But again, there was a separate use - a distinction that makes some difference as to whether the amount and substantiality of the copying was "reasonable in relation to the purpose of the copying" for the library copies. Campbell, 510 U.S. at 586.

(i) The Purchased Library Copies Converted from Print to Digital.

For the print library copies that Anthropic purchased and then converted into digital library copies, Anthropic already enjoyed entitlement to keep the copies in its library. The purpose of the copying was to keep them in its library but with more favorable storage and searchability properties. Copying the entire work was exactly what this purpose required. There was no surplus copying. The source copy was destroyed.

The third fair use factor favors fair use for the purchased library copies converted from print to digital.

(ii) The Pirated Library Copies.

For the pirated library copies, however, Anthropic lacked any entitlement to hold copies of the books at all. Its purpose, it says, was to train LLMs. But its objective conduct was to seek "all the books in the world" and then retain them even after deciding it would not make further copies from them for training - indicating there were other further uses. Against the purpose of acquiring all the books one could on the chance some might prove useful for training LLMs and maybe other stuff too, almost any unauthorized copying would have been too much. Anthropic copied millions of books in toto, Authors' works among them.

The third factor points against fair use for the pirated library copies.

4. THE EFFECT OF THE USE UPON THE MARKET FOR OR VALUE OF THE COPYRIGHTED WORK.

...

A. THE COPIES USED TO TRAIN SPECIFIC LLMS.

The copies used to train specific LLMs did not and will not displace demand for copies of Authors' works, or not in the way that counts under the Copyright Act.

...

The fourth factor thus favors fair use for the training copies.

B. THE COPIES USED TO BUILD A CENTRAL LIBRARY.

(i) The Purchased Library Copies Converted from Print to Digital.

... It was a format change.

...

This factor is thus neutral for the purchased library copies converted from print to digital.

(ii) The Pirated Library Copies.

The copies used to build a central library and that were obtained from pirated sources plainly displaced demand for Authors' books - copy for copy. ...

The fourth factor points against fair use for the pirated library copies.

5. OVERALL ANALYSIS.

After the four factors and any others deemed relevant are "explored, [ ] the results [are] weighed together, in light of the purposes of copyright." Campbell, 510 U.S. at 578.

The copies used to train specific LLMs were justified as a fair use. Every factor but the nature of the copyrighted work favors this result. The technology at issue was among the most transformative many of us will see in our lifetimes.

The copies used to convert purchased print library copies into digital library copies were justified, too, though for a different fair use. The first factor strongly favors this result, and the third favors it, too. The fourth is neutral. Only the second slightly disfavors it. On balance, as the purchased print copy was destroyed and its digital replacement not redistributed, this was a fair use.

The downloaded pirated copies used to build a central library were not justified by a fair use. Every factor points against fair use. Anthropic employees said copies of works (pirated ones, too) would be retained "forever" for "general purpose" even after Anthropic determined they would never be used for training LLMs. A separate justification was required for each use. None is even offered here except for Anthropic's pocketbook and convenience.

And, as for any copies made from central library copies but not used for training, this order does not grant summary judgment for Anthropic. On this record in this posture, the central library copies were retained even when no longer serving as sources for training copies, "hundreds of engineers" could access them to make copies for other uses, and engineers did make other copies. Anthropic has dodged discovery on these points (e.g., Opp. Exh. 17 at 93-94 (retained); Opp. Exh. 22 at 196 (no limits); Opp. Exh. 30 at 3, 4 (no accounting); see also Opp. 15). We cannot determine the right answer concerning such copies because the record is too poorly developed as to them. Anthropic is not entitled to an order blessing all copying "that Anthropic has ever made after obtaining the data," to use its words (Opp. Exh. 30 at 3, 4).

...

1.利用の目的及び性質

(略)

A.特定の大規模言語モデルの学習のために利用された複製

(略)

私たちの事実により近いものは-判決、意見書、法的見解の記事及び同様のものを用い-法的プロンプトを受け取り、新たな法的文章で答える学習したAIツールであろう。そして、これらに大いに似た事実に基づき、別の裁判所が異なる方向を取っている。それはフェアユースと判示したのである。2014年7月3日のホワイト対ウェスト出版社事件ニューヨーク南部地裁判決。

後者の利用は著作権者が正しくコントロールできると期待できるものと十分「直交する」。2023年5月18日のアンディ・ウォーホル財団対ゴールドスミス事件アメリカ最高裁判決参照。すなわち、それは「創作するインセンティブを減らす事なく科学と芸術の進歩を促進する」ものであって、著作権に縛られる事なく自由に利用できる。同上;アメリカ憲法第1条第8節第8項参照。

要するに、新しい文章を生成するために著作物を大規模言語モデルの学習に用いる事の目的及び性質は真に変形的なものである。作者となろうとする読者の様に、アンスロピックの大規模言語モデルは-作品の先を走りそれを再現するか置き換えるためにではなく-急な角を曲がり、何かしら別のものを作り出すために、作品に基づく学習を行ったのである。この学習プロセスにおいて大規模言語モデル等の中で複製がなされる事が合理的に必要であったのであれば、これらの複製は変形的利用においてなされたものである。

第1の要素は学習のための複製についてフェアユースに有利である。

B.中央ライブラリを構築するために用いられた複製

しかし、問題となっている利用はそれだけではない。アンスロピックがその中央ライブラリのために何百万冊もの書籍を購入し、同じくその中央ライブラリのために何百万ものデジタル本を違法にコピーした事を思い出さなくてはならない。それは特定の大規模言語モデルの学習のために特定の本の集合及び部分集合を利用した。そして、大規模言語モデルの学習のために(全く又は二度と)利用しないと決めた後でも、それは後で発生するかも知れない他の利用のためにその中央ライブラリにおける全ての複製を保持した。複製した作品の幾許かは大規模言語モデルの学習のために間々使われるため、それ以上の説明は要らず、アンスロピックは世界のあらゆる作品を自由に取り、永遠に保持しておく資格があるとアンスロピックは考えている様である。しかしながら、AI企業のための著作権法からの除外はない。

アンスロピックが購入し、違法にコピーしたライブラリの複製の間で法的問題は異なるため、本決定ではこれらを順番に検討する。

(ⅰ)紙からデジタルに変換された、購入されたライブラリの複製

(略)

まとめると、第1のフェアユースの要素は、購入された紙のライブラリの複製から変換されたデジタルのライブラリの複製にとって有利なものである-ただし、これらは違法にコピーされたライブラリの複製の口実となるものではない。

(ⅱ)違法にコピーされたライブラリの複製

(略)

まとめると、第1の要素において、違法なソースから作られた中央ライブラリの複製はフェアユースに反する方向にある。

2.著作物の性質

(略)

ここで、アンスロピックは全ての著作者の本-フィクションかノンフィクションのいずれにせよ、全て出版されたもの-は表現的要素を含んでいると認めている。そして、上で記した通り、本決定は、中央ライブラリを構築して特定の大規模言語モデルを学習させる上でその作品はその表現的性質のために選ばれたという著作者の見解を受け入れる。

(略)

第2の要素において、同種の全ての複製はフェアユースに反する方向にある。

3.利用された部分の量と本質性

(略)

A.特定の大規模言語モデルの学習のために利用された複製

(略)

「問題となるのは『利用された部分の量及び本質性』というより、[第1の利用の]代替として競合するものとなり得る、[主張された第2の利用において]公衆にアクセス可能とされたものの量及び本質である。」2015年10月16日の全米作家協会対グーグル社事件第2巡回控訴裁判決。ここで、また、クロードのサービスの出力と著作者の作品の間に何か追跡可能な繋がりがあるは主張されていない。すなわち、クロードの基礎にある大規模言語モデルの学習のために利用された複製は特に合理的なものである。

(略)

1つ、アンスロピックが所与の大規模言語モデルの学習のために何十億もの語が必要であった事は皆認めている。本だけを使うとして、アンスロピックはモデル毎に何百万冊もの本を必要としたであろう。本の小さな部分及び他の文章のより大きな部分を含む集合を使うとして、アンスロピックはなお何万冊もの本を必要としたであろう。アンスロピックはその様な本の小さな集合を利用できた事を示したのであるから、本を全く使わない-少なくとも彼らの本を使わない-事もできたと著作者は主張している。しかし、著作者は「合理的に必要」とは「厳密に必要」を意味しないという事を忘れている。著作者は大規模言語モデルの学習のために求められる文章の量が莫大である事を争っていない。それほど沢山の作品の利用が合理的に必要なのであるから、実際に大規模言語モデルの学習のためにある著作物を利用する事はその次のものと同じく合理的なものであろう。

もう1つ、公衆に対する出力による侵害は主張されてすらいない。よって、確かに、著作者の作品は文章の最も強い例として選ばれていた。しかし、強い例に基づいて大規模言語モデルを学習する事により当然得られる利益が、作品それ自体の部分の公衆への開示によって減殺されたという事もない。したがって、複製されたものは合理的であり、当然必要とされたものである。

よって、第3の要素は学習のための複製についてフェアユースに有利である。

B.中央ライブラリを構築するために利用された複製

しかし、また、別の利用-ライブラリの複製について、複製の量及び本質性が「複製の目的との関係で合理的」かどうかに関し、違いを生み出す区別-がある。1994年3月7日のキャンベル対エイカフ=ローズ・ミュージック社事件アメリカ最高裁判決。

(ⅰ)紙からデジタルに変換された、購入されたライブラリの複製

アンスロピックが購入してデジタルのライブラリの複製に変換した書籍のライブラリの複製について、アンスロピックは複製をそのライブラリに保持する資格を既に享受していた。複製の目的はそれをそのライブラリに保持する事であったが、より好ましい蓄積及び検索可能性を伴っての事である。作品全体の複製は正しくこの目的により求められる事である。追加の複製はない。元の複製は破棄された。

第3のフェアユースの要素は、紙からデジタルに変換された、購入されたライブラリの複製についてフェアユースに有利である。

(ⅱ)違法にコピーされたライブラリの複製

しかしながら、違法にコピーされたライブラリの複製について、アンスロピックは本の複製を保持する資格を全く持たない。それが言うには、その目的は大規模言語モデルの学習のためである。しかし、その客観的な振舞は、「世界における全ての本」を探して、そこから学習のためにさらなる複製を作らないと決定した後でも-他のさらなる利用がある事を示唆し-それを保持しようとしたというものである。全ての本を手に入れるという目的に対して、上手くすれば幾許かが大規模言語モデルの学習のために有益である事を証明できるかも知れないし、他のものについてもそうかも知れないが、ほとんどの違法な複製は行き過ぎであろう。アンスロピックは何百万もの本をすっかり複製しており、著作者の作品もその中にある。

第3の要素において、違法にコピーされたライブラリの複製はフェアユースに反する方向にある。

4.著作物の市場又はその価値に与える利用の影響

(略)

A.特定の大規模言語モデルの学習のために用いられた複製

特定の大規模言語モデルの学習のために用いられた複製が著作者の作品の複製に対する要求を置き換えたという事も、そうなるだろうという事も、著作権法の下で考慮されるやり方でそうなるという事もない。

(略)

よって、第4の要素は学習のための複製についてフェアユースに有利である。

B.中央ライブラリを構築するために利用された複製

(ⅰ)紙からデジタルに変換された、購入されたライブラリの複製

(略)それはフォーマットチェンジである。

(略)

よって、この要素は、紙からデジタルに変換された、購入されたライブラリの複製について中立である。

(ⅱ)違法にコピーされたライブラリの複製

中央ライブラリを構築されるために用いられた、違法ソースから手に入れた複製は-複製に対する複製であって-著作者の本への要求を明らかに置き換えるものである。

第4の要素において、違法にコピーされたライブラリの複製はフェアユースに反する方向にある。

5.総合分析

4つの要素及び関係するとみなされる他の事が「探求された後で、著作権の目的に照らして、結果はともに衡量される。」キャンベル事件アメリカ最高裁判決。

特定の大規模言語モデルの学習のために利用された複製はフェアユースとして正当化される。著作物の性質以外の全ての要素がこの結果に有利である。問題となる技術は私たちの多くが自分たちの人生において見るだろうものの中で最も変形的なものの1つである。

購入された書籍の複製をデジタルライブラリの複製に変換するために利用された複製もまた別にフェアユースとして正当化される。第1の要素はこの結果に強く有利なものであり、第3の要素も同じく有利である。第4の要素は中立である。第2の要素のみが若干それに不利である。比較衡量において、購入された紙の複製が破棄され、そのデジタルの置換物が再配布されていない以上、これはフェアユースである。

中央ライブラリを構築するために利用された、違法にダウンロードされた複製はフェアユースによって正当化されない。全ての要素においてこれはフェアユースに反する方向にある。作品の複製は(違法にコピーされたものも)、アンスロピックが大規模言語モデルの学習のために利用しないと決めた後でも「一般的な目的」のために「永遠に」保持されるであろうとアンスロピックの従業員は言った。それぞれの目的のために別の正当化事由が求められる。アンスロピックの手控えと利便を除き、ここで如何なる正当化事由も提示されていない。

そして、中央ライブラリの複製から作られたが学習のために利用されなかった複製について、本決定はアンスロピックに有利な略式判決を与えるものではない。この点に関する記録において、中央ライブラリの複製は学習のための複製の元としてもはや利用されなくなった時にも保持されており、「何百人ものエンジニア」が他の利用のためにアクセスして複製を作る事ができ、エンジニアたちは他の複製を作った。アンスロピックはこれらの点でディスカバリーを避けた。記録はこれらに関しては余りにも不十分であり、私たちはこの様な複製に関して正しい答えを決定する事ができない。アンスロピックは、その語を利用するために「データを手に入れた後でアンスロピックが作った」全ての複製を祝福する様な決定を受ける資格はない、

(略)

 上で書いた事の繰り返しになるが、この判決は、新しい文章を生成するために著作物を大規模言語モデルの学習に用いる事は変形的なものであって、多くの本を利用する事も合理的に必要な事であり、基本的にフェアユースと考えられるとしながらも、元の著作物の性質を見るだけの第2の要素を除き、第1、第3、第4の要素で、複製が合法か違法かによって判断を分け、適法に購入した本からした複製の利用であればフェアユースだが、違法コピーの利用はフェアユースに反するとしている。

(3)2025年6月25日のリチャード・カドレー他対メタ社事件カリフォルニア北部地裁判決

 今回最後に取り上げる判決(pdf)は、同じくカリフォルニア北部地方裁から、上のもののすぐ後の2025年6月25日にメタの生成AIに対して出されたものであり、以下の様に、新しい文章を作るための生成AIの学習への著作物の利用は基本的にフェアユースたり得るとしたものである。

III. FACTOR ONE: THE PURPOSE AND CHARACTER OF THE USE

...

This factor favors Meta. There is no serious question that Meta's use of the plaintiffs' books had a "further purpose" and "different character" than the books - that it was highly transformative. The purpose of Meta's copying was to train its LLMs, which are innovative tools that can be used to generate diverse text and perform a wide range of functions. Cf. Oracle, 593 U.S. at 30 (transformative to use copyrighted computer code "to create a new platform that could be readily used by programmers"). Users can ask Llama to edit an email they have written, translate an excerpt from or into a foreign language, write a skit based on a hypothetical scenario, or do any number of other tasks. The purpose of the plaintiffs' books, by contrast, is to be read for entertainment or education.

...

IV. FACTOR TWO: THE NATURE OF THE COPYRIGHTED WORK

...

This factor favors the plaintiffs. Their books - mostly novels, memoirs, and plays - are highly expressive works "of the type that the copyright laws value and seek to protect." Hachette, 115 F.4th at 187 (quoting Authors Guild, Inc. v. HathiTrust, 755 F.3d 87, 98 (2d Cir. 2014)). That some of their works may be factual (like an autobiography) as opposed to fictional does not meaningfully change this conclusion, because copyright still protects an author's "manner of expressing" facts. Google Books, 804 F.3d at 220.

...

V. FACTOR THREE: THE AMOUNT AND SUBSTANTIALITY OF THE PORTION USED IN RELATION TO THE COPYRIGHTED WORK AS A WHOLE

...

In any event, this factor favors Meta, even though it copied the plaintiffs' books in their entirety. The amount that Meta copied was reasonable given its relationship to Meta's transformative purpose. See Oracle, 593 U.S. at 34. Everyone agrees that LLMs work better if they are trained on more high-quality material. See Ungar Decl. ISO Meta MSJ ¶¶ 42-48; Pls. Reply Ex. 115 ¶¶ 79-80. So feeding a whole book to an LLM does more to train it than would feeding it only half of that book. With this in mind, it was "reasonably necessary" for Meta to "make use of the entirety of the works." HathiTrust, 755 F.3d at 98.

VI. FACTOR FOUR: THE EFFECT OF THE USE UPON THE POTENTIAL MARKET FOR OR VALUE OF THE COPYRIGHTED WORK

...

In a case involving the use of copyrighted works to train generative AI models, there are at least three ways a plaintiff might try to argue that the defendant's copying harmed the market for the works (or that the market would be harmed if that copying were widespread). First, the plaintiff might claim that the model will regurgitate their works (or outputs that are substantially similar), thereby allowing users to access those works or substitutes for them for free via the model. Second, the plaintiff might point to the market for licensing their works for AI training and contend that unauthorized copying for training harms that market (or precludes the development of that market). Third, the plaintiff might argue that, even if the model can't regurgitate their own works or generate substantially similar ones, it can generate works that are similar enough (in subject matter or genre) that they will compete with the originals and thereby indirectly substitute for them. In this case, the first two arguments fail. The third argument is far more promising, but the plaintiffs' presentation is so weak that it does not move the needle, or even raise a dispute of fact sufficient to defeat summary judgment.

A

If Llama could be used to generate significant portions of the plaintiffs' books - or text so similar to their books as to be infringing in its own right - that would threaten the market for the books because people would read those outputs instead. But that theory of harm is not viable in this particular case because, as discussed above, Llama does not allow users to generate any meaningful portion of the plaintiffs' books. Neither party's expert opined that Llama was able to regurgitate more than 50 words from any of the plaintiffs' books, even in response to "adversarial" prompting designed specifically to make LLMs regurgitate. ...

B

The plaintiffs' primary theory of market harm is that Meta's unauthorized use of their books for LLM training harms the market for licensing their books for that purpose. The plaintiffs devote nearly all of their discussion of the fourth factor to this theory. The parties therefore go back and forth at length about whether a market for licensing general trade books exists or is likely to develop.

But whether such a market exists or is likely to develop is irrelevant, because this market is not one that the plaintiffs are legally entitled to monopolize. In every fair use case, the "plaintiff suffers a loss of a potential market if that potential [market] is defined as the theoretical market for licensing" the use at issue in the case. Tresona Multimedia, LLC v. Burbank High School Vocal Music Association, 953 F.3d 638, 652 (9th Cir. 2020) (emphasis omitted) (quoting 4 Melville B. Nimmer & David Nimmer, Nimmer on Copyright § 13.05 (2019)). Therefore, to prevent the fourth factor analysis from becoming circular and favoring the rightsholder in every case, harm from the loss of fees paid to license a work for a transformative purpose is not cognizable. Id.; Bill Graham Archives v. Dorling Kindersley Ltd., 448 F.3d 605, 614-15 (2d Cir. 2006); see also Oracle, 593 U.S. at 38 ("cautioning against the 'danger of circularity'" (quoting 4 Nimmer § 13.05)).

C

The third way that using copyrighted books to train an LLM might harm the market for those works is by helping to enable the rapid generation of countless works that compete with the originals, even if those works aren't themselves infringing. Assume for this discussion that people can (or will soon be able to) use LLMs to generate massive amounts of text in significantly less time than it would take to write that text, and using a fraction of the creativity. People could thus use LLMs to create books and then sell them, competing with books written by human authors for sales and attention. Indeed, to some extent, this appears to be occurring already - one expert for the plaintiffs briefly discusses reports of AI-generated books "flooding Amazon." Pls. MSJ Ex. 76 ¶ 199; see id. ¶¶ 193-207. People might even be motivated to make those books available for free, given how easily it will presumably be to prompt an LLM to create them. Harm from this form of competition is the harm of market dilution. Or as one commentator describes it, the harm of "indirect" substitution, rather than "direct" substitution (which would be the first form of harm described). See Matthew Sag, Fairness and Fair Use in Generative AI, 92 Fordham L. Rev. 1887, 1916-20 (2024).

Of course, not all copyrighted works would have their markets diluted equally by AI-generated competitors. It seems unlikely, for instance, that AI-generated books would meaningfully siphon sales away from well-known authors who sell books to people looking for books by those particular authors. But it's easy to imagine that AI-generated books could successfully crowd out lesser-known works or works by up-and-coming authors. While AI-generated books probably wouldn't have much of an effect on the market for the works of Agatha Christie, they could very well prevent the next Agatha Christie from getting noticed or selling enough books to keep writing.

This effect also seems likely to be more pronounced with respect to certain types of works. For instance, an AI model that can generate high-quality images at will might be expected to greatly affect the market for such images, diminishing the incentive for humans to create them. An LLM that could generate accurate information about current events might be expected to greatly harm the print news market. The market for certain nonfiction works - for example, books about how to take care of your garden - could be greatly diminished by the ability of LLMs to produce books on that topic. For fiction works, it might be more dependent on the author or the genre in which that author operates.

... Meta introduced evidence that its copying hasn't caused market harm. The plaintiffs presented no empirical evidence to the contrary - no evidence that the copying has already caused market harm, and no evidence that the copying is likely to cause market harm in the future. All the plaintiffs presented is speculation, and speculation is insufficient to raise a genuine issue of fact and defeat summary judgment. E.g., Anheuser-Busch, Inc. v. Natural Beverage Distributors, 69 F.3d 337, 345 (9th Cir. 1995).

... Nor is it self-evident that Llama will harm the book sale market by enabling users to create a flood of competing books. It's possible, even likely, that Llama will harm the book sale market. But to conclude that it will requires inferring that Llama (and not just any LLM) can be used to create such books, that it will be used to create such books, that consumers will purchase those books instead of books written by human authors, that consumers will buy those books instead of the plaintiffs' books in particular, and that Llama is meaningfully better at creating those books because it was trained on copyrighted material. ...

VII. CONCLUSION

...

In this case, because Meta's use of the works of these thirteen authors is highly transformative, the plaintiffs needed to win decisively on the fourth factor to win on fair use. See, e.g., Perfect 10, 508 F.3d at 1168 (fair use where secondary use was "significant[ly] transformative" and fourth factor "favor[ed] neither party"). And to stave off summary judgment, they needed to create a genuine issue of material fact as to that factor. Because the issue of market dilution is so important in this context, had the plaintiffs presented any evidence that a jury could use to find in their favor on the issue, factor four would have needed to go to a jury. Or perhaps the plaintiffs could even have made a strong enough showing to win on the fair use issue at summary judgment. But the plaintiffs presented no meaningful evidence on market dilution at all. Absent such evidence and in light of Meta's evidence, the fourth factor can only favor Meta. Therefore, on this record, Meta is entitled to summary judgment on its fair use defense to the claim that copying these plaintiffs' books for use as LLM training data was infringement.

...

Ⅲ.要素1:利用の目的及び性質

(略)

この要素はメタに有利である。原告の本のメタによる利用が本とは別の「さらなる目的」及び「異なる性質」を持っていた事にほぼ疑問の余地はない-これは高度に変形的なものである。メタの複製行為の目的は、様々な文章を生成し、広い範囲の機能を発揮する事に用いられ得る革新的なツールであるその大規模言語モデルの学習にある。2021年4月5日のグーグル社対米オラクル社事件アメリカ最高裁判決(「プログラマーによってすぐに利用できる新しいプラットフォームを生成するため」著作権の保護を受けるコンピューターコードを変形的に利用するもの)。利用者はLlamaに、書いた電子メールを添削する事、抜粋の外国語から又は外国語に翻訳する事、仮のあらすじに基づき寸劇を書く事、又は他にも数多くの仕事をする事を頼む事ができる。原告の本の目的は、それに対し、娯楽又は教育のために読まれる事にある。

(略)

Ⅳ.要素2:著作物の性質

(略)

この要素は原告に有利である。その本は-ほとんど小説、回想録及び戯曲である-は、「著作権法が評価し、保護しようとする種類の」高度に表現的な作品である。2024年9月4日のアシェット・ブック・グループ社対インターネット・アーカイブ事件第2巡回控訴裁判決(2015年10月16日の全米作家協会対グーグル社事件第2巡回控訴裁判決を引用)。著作権が保護するのは著作者が事実を「表現するやり方」であるから、その作品の幾つかがフィクションではなく(自伝の様な)事実を書いたものである事がこの結論を変える事はない。2015年10月16日の全米作家協会対グーグル社事件第2巡回控訴裁判決。

(略)

Ⅴ.要素3:著作物全体との関係で用いられた部分の量及び本質性

(略)

いずれにしても、メタが原告の本全体を複製したにせよ、この要素はメタに有利である。メタが複製した量はメタの変形的利用との関係の下で合理的である。オラクル社事件アメリカ最高裁判決参照。より多くの高品質なマテリアルによって学習する時、大規模言語モデルがより良く動く事に皆同意している。意見書等参照。よって、本全体を大規模言語モデルに与える事は、本の半分のみを与えるより、学習に対して多くをなす。この事を踏まえ、メタが作品の全体を利用する事は「合理的に必要」なものであった。2014年6月10日の全米作家協会対ハイチ・トラスト事件第2巡回控訴裁判決。

Ⅵ.要素4:著作物の市場又はその価値に与える利用の影響

(略)

生成AIモデルの学習のための著作物の利用が関係する事件において、被告の複製行為がその著作物の市場を害した(又はその複製行為が広がるとその市場が害されるであろう)と原告が主張するやり方は少なくとも3つある。第1に、原告は、モデルがその作品をその吐き出し(又は実質的に類似したものを出力し)、それによってそのモデルを通じてその作品又はそれに対する代替物を利用者にアクセス可能としたと主張できるだろう。第2に、原告は、その作品をAI学習のためにライセンスする市場を示し、許諾を得ていない学習のための複製行為がその市場を害する(又はその市場の発展を阻害する)と主張できるだろう。第3に、原告は、そのモデルが彼ら自身の作品をその儘吐き出すか実質的に類似するものを生成し得ないとしても、それは、元の作品と競合し、それによって間接的にその代替となるであろう、(題材又はジャンルにおいて)十分類似した作品を生成する事ができると主張できるだろう。本事件において、最初の2つは失敗する。第3のものは大いに見込みがあるが、原告による提示は非常に弱く、針を動かさないか、略式判決を退けるのに十分なだけの事実の争いを提起しない。

もしLlamaが原告の本の大きな部分-又は彼ら自身の権利の侵害となるほどその本に類似した文章-を生成するのに使われ得るようであれば、人々は代わりにその出力を読むであろうから、その本の市場を脅かす事になるであろう。しかし、上で議論した通り、Llamaは原告の本の意味のある部分を生成する事を利用者に許していないのであるから、この害に関する理論はこの個別のケースにおいて有効ではない。大規模言語モデルにその儘吐き出させるために特別に設計された「敵対」プロンプト入力に対する応答としても、いずれの当事者の専門家もLlamaは原告の本のいずれからも50語もその儘吐き出す事はできないと述べている。(略)

市場への害に関する原告の主たる理論は、メタの大規模言語モデルの学習のためのその本の無許諾利用がその目的についてその本をライセンスする市場を害するというものである。原告は第4の要素に関するその議論のほぼ全てをこの理論にささげている。そのため、当事者は一般取引本のライセンス市場が存在するか、発展しようとしているかについて長々と行ったり来たりを繰り返している。

しかし、この市場は原告が法的に独占する資格があるものではないため、その様な市場が存在するか、発展しようとしているかは無関係である。フェアユースが関係するあらゆる事件において、「もし潜在的なもの[市場]が」その事件において問題となっている利用に対する「理論的なライセンス市場であると定義されるなら、原告は潜在市場の喪失を蒙る」。2020年3月24日のトレゾナ・マルチメディア社対バーバンク高校声楽連盟事件第9巡回控訴裁判決(ニンマー著作権法を引用)。したがって、第4の要素の分析が循環的なものとなり、あらゆる事件において権利者に有利なものとなるのを避けるため、変形的な目的について作品をライセンスするために支払われるべき料金の損失から来る害は認められない。同上;2006年5月9日のビル・グラハム・アーカイブズ対ドーリング・キンダースレイ社事件第2巡回控訴裁判決;オラクル社事件アメリカ最高裁判決(「『循環性の危険』に注意すること」(ニンマー著作権法を引用))も参照。

大規模言語モデルの学習に著作権の保護を受ける本を利用する事がその作品の市場を害し得るとする第3のやり方は、その作品自体は侵害するものでないとしても、元の著作物と競合する無数の作品を急速に生成する事を可能にする事を助けるという事による。この議論は、創造性の僅かな欠片を利用し、人々が大規模言語モデルを用いてその文章を書くのにかかるより非常に少ない時間で大量の文章を生成する事ができる(又はすぐにできる様になる)という事を前提とする。そうすると、人々は大規模言語モデルを用いて本を作り出して売る事ができ、販売のためと耳目を集めるために人間の著作者によって書かれた本と競合する事になるだろう。実際、ある程度まで、この事は既に置きている様であり-ある専門家は原告のために「アマゾンに溢れる」AI生成本の報告について簡単に議論している。原告意見書。大規模言語モデルにプロンプトを入力してそれを作る事がどれほど簡単になると想定されるかを踏まえると、人々はこれらの本を無料で入手可能とする気にまでなるかも知れない。この形の競争からくる害は市場の希釈化の害である。そして、ある論文はそれを(説明される第1の形の害であろう)「直接的」代替というより「間接的」代替の害であると説明している。マシュー・サグ著、生成AIにおける公正性及びフェアユース(2024年)参照。

もちろん、全ての著作物がその市場においてAI生成の競争者によって等しく希釈化される訳ではないであろう。例えば、AI生成本がその個別の著者による本を探し求める人々に本を売る良く知られた著作者から売上を大きく吸い上げる事はありそうにないと思える。しかし、AI生成本がそれほど知られていない作品や新進の著作者による本を上手く押し出し得ると想像するのは容易い。AI本は恐らくアガサ・クリスティーの作品の市場に対してそれほど影響を与えないであろうが、それは次のアガサ・クリスティーが気づかれ、書き続けられるほど十分に売れる事を非常に良く妨げ得るであろう。

また、この影響は特定の種類の作品に関してより顕著なものになるだろうと思える。例えば、高品質な画像を好きに生成できるAIモデルはそのような画像の市場に大いに影響を与える事が予想され、それを作ろうとする人間のインセンティブを減退させるであろう。現在の出来事に関する正確な情報を生成できる大規模言語モデルは刊行ニュース市場を大いに害する事が予想される。何かしらのノンフィクション作品-例えば自身の庭を手入れする方法に関する本-の市場はそのトピックに関する本を作成する大規模言語モデルの能力によって大いに減少し得るであろう。フィクション作品については、著者及びその著者が活動するジャンルにより依存するものだろう。

(略)メタはその複製行為が市場に害をもたらしていない証拠を提出した。原告はそれに反対するものとして実証的証拠を-複製行為が既に市場に害をもたらしている証拠も、複製行為が将来的に市場に害をもたらすであろう事の証拠も-何ら提示していない。原告が提示したのは憶測に過ぎず、憶測は、事実に関する本質的な問題を提起し、略式判決を退けるには不十分である。例えば、1995年10月27日のアンホイザー・ブッシュ社対ナチュラル・ビバレッジ・ディストリビューターズ事件第9巡回控訴裁判決。

(略)Llamaが利用者に競合する本の洪水の作成を可能とする事によって本の販売市場を害するという事も自明ではない。Llamaが本の販売市場を今後害する事はあり得、そうなりそうでもある。しかし、そう結論するには、(何らかの大規模言語モデルとしてのみではなく)Llamaがその様な本を作り出す事に使われ得る事、その様な本を作り出す事に使われるだろう事、消費者が人間の著作者によって書かれた本の代わりにその本を購入するだろう事、消費者が特に原告の本の代わりにその本を買うだろう事、そして、著作権の保護を受けるマテリアルによって学習したためにその本を大いにより良く作る事の推論が必要だろう。(略)

Ⅶ.結論

(略)

本事件においては、メタによるこれらの十三人の著作者の作品の利用は高度に変形的なものであるから、原告はフェアユースにおいて勝つためには第4の要素で決定的に勝つ必要があった。例えば、2007年5月16日のパーフェクト10社対アマゾン・ドットコム社事件第9巡回控訴裁判決(二次利用は「とても変形的なもの」であり、第4の要素は「いずれの当事者にも有利なものではなかった」場合におけるフェアユース)参照。そして、略式判決を免れるためには、彼らはこの要素に関して具体的な事実に関する本質的な問題を作り出す必要があった。市場の希釈化の問題はこの文脈において実に重要であり、陪審がこの問題に関して原告に有利となる事実の認定に使える証拠を提示したなら、要素4は陪審に送られる必要があったであろう。あるいは原告は略式判決においてフェアユースの問題について勝てるだけ十分強いものを見せる事ができたかも知れない。しかし、原告は市場の希釈化について有意な証拠を何ら示していない。その様な証拠の欠如及びメタの証拠に照らし、第4の要素はメタのみに有利である。したがって、本記録に基づき、これらの原告の本を大規模言語モデルの学習データとして使う事は侵害であるとする主張に対するそのフェアユースの抗弁についてメタは略式判決を受ける資格を有する。

(略)

 上で書いた通り、この判決も、新しい文章を作りための生成AIの学習への著作物の利用は変形的なものであり、その様な生成AIを開発するために多くの本の全体を利用する事も合理的に必要な事であって、元の著作物の市場への影響も証明できておらず、基本的にフェアユースに該当するとしている。

 そして、この判決の特徴は、第4の要素の考察において、市場への影響として3つの場合が考えられるとして細かな分析を加えている事にあると言って良いだろう。

 1つ目は、生成AIが元の著作物をそのまま出力する事によって市場における代替が生じる場合、2つ目のものは生成AIへの学習に対するライセンス市場に対する影響を考える場合である。第1の場合については、この判決で述べられている通り、元の著作物をそのまま出力する事が技術的に止められている場合には問題とならないだろう。また、第2の場合についても、本来あの利用が公正なものであってライセンス不要かどうかが議論されるべきであるのに、その様な利用についてライセンスしていたかできたかも知れないといったライセンス市場への影響に関する主張を認めると、如何なる場合でも市場への影響があるというおかしな事になるので、フェアユース分析においてライセンス市場を問題とするべきではないのは理論的に当然の事と言って良い。なお、日本ではまだ著作権法第30条の4の非享受目的利用に関する判決が出されているという事はないが、そのただし書きについてもここで言われている事と同様に考えるべきである。

 この判決が取り上げている中で最も議論を呼ぶのが3つ目の、生成AIによって直接的に侵害となるものではないが同種のものが大量に生成される事で市場が希釈化・ダイリューションされる(間接的に代替される)とする様な場合だろう。

 十分な証拠がないと言いながら、最も見込みがあると言って長々と講釈を垂れているあたり、この地裁の判事はこの市場の希釈化の理論にかなりの愛着を見せているように見えるが、果たして著作権法に基づくフェアユースの世界にこの様な市場の希釈化の概念を持ち込んで良いのか私には甚だ疑問に思える。

 生成AIが元の著作物とは別のものだが同種のものを大量に作り出す事により、消費者が人間の著作者によって作成されたものの代わりにAI生成物を購入し、結果として著作物の市場に害を与える事自体ありそうにないが、たとえその様な事が一部で起こったとしても、大体、写真と絵の関係1つ取っても明らかな様に、この様な間接的な代替自体は今までも技術の発展において起こって来た事に過ぎないのであって、最近のAI技術だけ特別視するべき理由は今の所ないと私は考えている。ここで、AI生成物が直接侵害物となる様な場合は別に考えるべきである事には注意が必要であり、また、それもまだ今の所ありそうにないが、このブログで以前から書いている通り、AI本当に人間の知的創作を代替する様になったとしたら、それはもはや学習への著作物の利用がどうとかいうレベルの問題ではなく創作保護法としての知的財産法の根底から考え直すべき話である。

 上で書いた通り、これらの判決はいずれもまだ地裁レベルの略式判決であって、これで何かが決まるという事はないと言って良いが、AI学習への著作物の利用のフェアユース該当性の判断において論点となり得りそうな事はほぼ出尽くしていると見て良いのではないかと思う。最初の(1)の事件は生成AIを扱っていない事に留意が必要であり、いずれの要素も相互に関係する所もあるが、最後に生成AIの学習への著作物の利用とフェアユースに関する論点を各要素毎になるべく簡単にまとめて書いておくと、

  • 第1の要素(利用の目的及び性質)において、生成AIの出力が十分変形的と考えられるかどうか、言い換えると、学習した元の著作物をその儘出力できる様になっていないか
  • 第2の要素(著作物の性質)において、著作物の創作性は高度なものか
  • 第3の要素(利用された著作物の部分の量と本質性)において、生成AIへの著作物の利用は合理的に必要か
  • 第4の要素(著作物の価値と市場への影響)において、市場の著作物を直接的に代替する事があるか、また、市場の希釈化(間接的な代替)をどの様に考えるか
  • 全体を通して、著作物を合法に入手したか、それとも違法コピーを利用したか、

となるだろう。

 さらに、第2の要素は大きな要素ではない事と、第3の要素で生成AIの開発に大量の著作物が必要である事に議論の余地はほぼない事を考えると、それぞれの事件で大きな論点になるのは、

  • 生成AIの出力の変形性(直接侵害物の出力の不可能化)
  • 違法コピーの使用の有無

の2点であり、今後の訴訟の展開次第で、

  • 市場の希釈化による影響の是非

も議論されて行く事になるのではないだろうか。

 最後、これらの様な事件はアメリカ最高裁まで行くのではないかと思うが、上でも書いた通り、アメリカでも、著作権法に基づくフェアユースの判断において市場の希釈化を持ち出すのは筋違いであるとされる事を、その結果として、新しい文章などを生成するAIの学習への著作物への利用は基本的にフェアユースだが、違法コピーをそうと知って使う様な場合は除くという合理的な判断がなされる事を私としては大いに期待している。そうすると、日本における著作権法第30条の4の非享受目的利用の権利制限の考え方と同じ形に収斂して行く事になるだろうが、著作権が本来守るべきものは何かという事を考えれば、それは極めて妥当な事と私には思えるのである。(私は完全なフェアユースとは思っていないが、日本版フェアユースとも言われる、著作権法第30条の4とAIの関係については、第492回で取り上げた文化庁の報告書参照。なお、アメリカの判決で言われた事などから、日本でも、著作権法第30条の4で市場の希釈化によって権利者の利益が不当に害されると言う者がいるかも知れないが、いずれの国でも、その様な間接的な代替による市場への影響は本来著作権法が止めるべき筋合いのものではないだろう。)

(2025年8月25日夜の追記:(3)の判決のリンク先が間違っていたのを直し、合わせて最後から2段落目の文章もその方が読み易いのではないかと箇条書きの形に書き直した。)

(2025年8月27日夜の追記:ロイターの記事などによると、(2)のアンスロピックの事件は当事者間で和解の合意が成立した様である。アンスロピックは合法コピーのAI学習への利用についてはフェアユースで勝ったが、違法コピーについてはこれ以上争っても不利と判断したのかも知れない。)

(2025年9月11日夜の追記:タイトルと1行目の変換ミスを直した(「人口」→「人工」)。)

| | コメント (0)

2025年7月27日 (日)

第515回:人工知能(AI)は著作者たり得ないとする2025年3月18日のアメリカ控訴裁の判決

 参院選の結果、自公与党が参院でも過半数を割り、政治情勢がますます流動化している中、日本の政策検討も停滞する可能性があると思うが、ここでは知財動向の話を続けて行く。

 人工知能(AI)と著作権の関係について、アメリカでは、著作物のAI学習への利用はファユースであり得るとする様な地裁判決も出されており、その内容についてもじきに見て行きたいと思っているが、まず、今まで書いてきた事の良い補足となるだろう、2025年3月18日のコロンビア特別区巡回控訴裁によるターラー対パールムッター事件判決の事を先に取り上げる。これは、第483回で取り上げた2023年8月18日のコロンビア地裁の判決の後、控訴裁も同様に、AIは著作者たり得ず、AIを著作者とする著作権登録は認められないとしたアメリカ著作権局の決定を妥当と判断したものである。(なお、言わずもがなかも知れないが、この事件の原告は、第511回で取り上げた日本のAI発明に関する事件も含め、第491回で取り上げたアメリカ特許庁のガイダンスと合わせて紹介した通り、多くの国でAIを発明者と認めるべきという訴訟を起こし、世界的大議論を惹起しているターラー氏その人である。)

 その判決(pdf)から、具体的な判断内容が書かれた部分を以下に訳出する。(以下、いつも通り翻訳は全て拙訳である。)

III

As a matter of statutory law, the Copyright Act requires all work to be authored in the first instance by a human being. Dr. Thaler's copyright registration application listed the Creativity Machine as the work's sole author, even though the Creativity Machine is not a human being. As a result, the Copyright Office appropriately denied Dr. Thaler's application.

A

Authors are at the center of the Copyright Act. A copyright "vests initially in the author or authors of the work." 17 U.S.C. § 201(a). And copyright protection only "subsists *** in original works of authorship[.]" Id. § 102(a).

The Copyright Act does not define the word "author." But traditional tools of statutory interpretation show that, within the meaning of the Copyright Act, "author" refers only to human beings. To start, the text of multiple provisions of the statute indicates that authors must be humans, not machines. In addition, the Copyright Office consistently interpreted the word author to mean a human prior to the Copyright Act's passage, and we infer that Congress adopted the agency's longstanding interpretation of the word "author" when it re-enacted that term in the 1976 Copyright Act.

Numerous Copyright Act provisions both identify authors as human beings and define "machines" as tools used by humans in the creative process rather than as creators themselves. Because many of the Copyright Act's provisions make sense only if an author is a human being, the best reading of the Copyright Act is that human authorship is required for registration.

First, the Copyright Act's ownership provision is premised on the author's legal capacity to hold property. A copyright "vests initially in the author[.]" 17 U.S.C. § 201(a). This means an "author gains 'exclusive rights' in her work immediately upon the work's creation." Fourth Estate Pub. Benefit Corp. v. Wall-Street.com, LLC, 586 U.S. 296, 300-301, (2019) (quoting 17 U.S.C. § 106). Because a copyright is fundamentally a property right created by Congress, and Congress specified that authors immediately own their copyrights, an entity that cannot own property cannot be an author under the statute.

Second, the Copyright Act limits the duration of a copyright to the author's lifespan or to a period that approximates how long a human might live. A copyright generally "endures for a term consisting of the life of the author and 70 years after the author's death." 17 U.S.C. § 302(a). The Copyright Office maintains "current records of information relating to the death of authors of copyrighted works" so that it can determine when copyrights expire. Id. § 302(d). If the author's death is unknown, the Copyright Act presumes death after "a period of 95 years from the year of first publication of a work, or a period of 120 years from the year of its creation[.]" Id. § 302(e). And even when a corporation owns a copyright under the work-made-for-hire provision, the copyright endures for the same amount of time - "95 years from the year of first publication" or "120 years from the year of its creation[.]" Id. § 302(c). Of course, machines do not have "lives" nor is the length of their operability generally measured in the same terms as a human life.

Third, the Copyright Act's inheritance provision states that, when an author dies, that person's "termination interest is owned, and may be exercised" by their "widow or widower," or their "surviving children or grandchildren," 17 U.S.C. § 203(a)(2), (A). Machines, needless to say, have no surviving spouses or heirs.

Fourth, copyright transfers require a signature. To transfer copyright ownership, there must be "an instrument of conveyance" that is "signed by the owner[.]" 17 U.S.C. § 204(a). Machines lack signatures, as well as the legal capacity to provide an authenticating signature.

Fifth, authors of unpublished works are protected regardless of the author's "nationality or domicile." 17 U.S.C. § 104(a). Machines do not have domiciles, nor do they have a national identity.

Sixth, authors have intentions. A joint work is one "prepared by two or more authors with the intention that their contributions be merged into inseparable or interdependent parts of a unitary whole." 17 U.S.C. § 101. Machines lack minds and do not intend anything.

Seventh, and by comparison, every time the Copyright Act discusses machines, the context indicates that machines are tools, not authors. For example, the Copyright Act defines a "computer program" as "a set of statements or instructions to be used directly or indirectly" to "bring about a certain result." 17 U.S.C. § 101. The word "machine" is given the same definition as the words "device" and "process," id., and those terms are consistently used in the statute as mechanisms that assist authors, rather than as authors themselves, id. §§ 102(a); 108(c)(2); 109(b)(1)(B)(i); 116(d)(1); 117(a)(1), (c); 401(a); 1001(2), (3). In addition, when computer programs and machines are referenced in the statute, the statute presumes they have an "owner," id. § 117(a), (c), who can perform "maintenance," "servic[e]," or "repair" on them, id. § 117(d)(1), (2).

All of these statutory provisions collectively identify an "author" as a human being. Machines do not have property, traditional human lifespans, family members, domiciles, nationalities, mentes reae, or signatures. By contrast, reading the Copyright Act to require human authorship comports with the statute's text, structure, and design because humans have all the attributes the Copyright Act treats authors as possessing. The human-authorship requirement, in short, eliminates the need to pound a square peg into a textual round hole by attributing unprecedented and mismatched meanings to common words in the Copyright Act. See Food & Drug Admin. v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 133 (2000) ("It is a 'fundamental canon of statutory construction that the words of a statute must be read in their context and with a view to their place in the overall statutory scheme.'") (quoting Davis v. Michigan Dept. of Treasury, 489 U.S. 803, 809 (1989)).

To be clear, we do not hold that any one of those statutory provisions states a necessary condition for someone to be the author of a copyrightable work. An author need not have children, nor a domicile, nor a conventional signature. Even the ability to own property has not always been required for copyright authorship. Married women in the nineteenth century authored work that was eligible for copyright protection even though coverture laws forbade them from owning copyrights. See Melissa Homestead, AMERICAN WOMEN AUTHORS AND LITERARY PROPERTY, 1822-1869, at 21-62 (2005); Belford, Clarke & Co. v. Scribner, 144 U.S. 488, 504 (1892) (recognizing Mrs. Terhune's authorship when her book's copyright was infringed, even though, as a married woman, she could not own property).

The point, instead, is that the current Copyright Act's text, taken as a whole, is best read as making humanity a necessary condition for authorship under the Copyright Act. That is the reading to which "the provisions of the whole law" point. John Hancock Mut. Life Ins. Co. v. Harris Tr. & Sav. Bank, 510 U.S. 86, 94 (1993) (quoting Pilot Life Ins. Co. v. Dedeaux, 481 U.S. 41, 51 (1987)).

制定法の問題として、著作権法は、あらゆる著作物は最初の段階で人間により著作される事を求めている。その創作機は人間でないにも関わらず、ターラー博士の著作権登録申請は唯一の著作者として創作機を挙げる。結果として、アメリカ著作権局はターラー博士の登録申請を適切に拒絶した。

著作者は著作権法の中心である。著作権は最初に著作物の著作者に付与される。アメリカ著作権法第201条(a)。そして、著作権保護が「あるのは著作者による独自の著作物に」のみである。同第102条(a)。

著作権法は「著作者」の語を定義していない。しかし、法律解釈の伝統的なやり方から、著作権法の意味における「著作者」は人間のみを指している事が分かる。まず、法規の幾つもの規定の文章が著作者は機械ではなく人間でなければならない事を示している。さらに、アメリカ著作権局は著作権法の文章に先立ち一貫して著作者の語を人間を意味するものと解釈して来ており、私たちは、1976年著作権法で再立法する際にアメリカ議会が著作権局の「著作者」の語の長きに渡る解釈を採用したと推察する。

数多くの著作権法の規定は、著作者を人間と定義するとともに、「機械」を創作者というよりむしろ創作過程において人間によって使われる道具と定義している。著作権法の規定の多くが著作者が人間である場合のみ意味をなすため、人間による著作者性が登録に求められるというのが著作権法の最も良い読み方である。

第1に、著作権法の所有規定は著作者の法的権利能力を前提としている。著作権は「最初に著作者に付与される。」著作権法第201条(a)。この事は「著作者が作品の創作において直ちにその作品に対する『排他的権利』を得る事を意味する。」フォース・エステート・パブリック・ベネフィット社対ウォール・ストリート・ドットコム社アメリカ事件2019年3月4日アメリカ最高裁判決(アメリカ著作権法第106条を引用)。著作権は本質的に議会によって作り出された権利であり、議会は著作者は直ちにその著作権を所有すると特定したのであるから、権利を所有できない存在は法規の下で著作者たり得ない。

第2に、著作権法は著作権の保護期間を著作者の寿命又は人間がおおよそどれくらい生きるだろうかという期間に制限している。著作権は一般的に「著作者の人生及び著作者の死後70年からなる期間存続する。」アメリカ著作権法第302条(a)。アメリカ著作権局は、いつ著作権が満了するのかを決められるよう、著作権の保護を受ける著作物の著者の死に関する情報の今現在の記録を保持している。同第302条(d)。著作者の死が不明の場合、著作権法は「著作物の最初の公開の年から95年の期間又はその創作の年から120年の期間」の後に死を推定する。同第302条(e)。そして、会社が職務著作規定の下で著作権を所有できるとしても、著作権は同じだけの期間-「著作物の最初の公開の年から95年の期間」又は「その創作の年から120年の期間」-存続する。もちろん、機械が「命」を持つ事はなく、それが機能する長さが人間の人生と同じ期間で一般的に測られる事もない。

第3に、著作権法の相続規定は、著作者が死んだとき、その者の「終了の利益は」その「寡婦又は寡夫」又はその「生存している子供又は孫」によって「所有され、行使され得る、」アメリカ著作権法第203条(a)(2)、(A)。機械は、言うまでもなく、生存している配偶者又は相続人を持たない。

第4に、著作権の移転は署名を必要とする。著作権の所有を移転するためには、「所有者によって署名された」「譲渡の文書」がなければならない。アメリカ著作権法第204条(a)。機械は署名できず、署名の認証を提供する法的能力も持たない。

第5に、未公開の著作物の著者は著者の「国籍又は居所」によらず保護される。アメリカ著作権法第104条(a)。機械は居所を持たず、国籍も持たない。

第6に、著作者は意図を有する。共同著作は「その寄与が一つの全体の不可分の又は互いに依存する部分に融合するという意図の下で複数の著作者により用意される」ものである。アメリカ著作権法第101条。機械に精神はなく、何かを意図する事もない。

第7に、また比較により、著作権法が機械について議論する場合はいつも、その文脈は機械が道具であって著作者ではない事を示している。例えば、著作権法は「コンピュータープログラム」を「ある結果をもたらす」ために「直接的に又は間接的に用いられる記述又は指示の組み合わせ」と定義している。アメリカ著作権法第101条。「機械」という語は「機器」や「方法」といった語と同様の定義を与えられており、同条、これらの用語は一貫して著作者そのものというよりむしろ著作者を補助する機構として法規中で用いられている、同第102条(a):第108条(c)(2);第109条(b)(1)(B)(ⅰ);第116条(d)(1);第117条(a)(1)、(c);第401条(a);第1001条(2)、(3)。さらに、コンピュータープログラム及び機械が法規において言及される場合は、法規は、それには「所有者」がおり、同第117条(a)、(c)、この者はその「維持」、「補修」又は「修理」を行える事を想定している、同第117条(d)(1)、(2)。

これらの法規の規定は全て集合的に「著作者」は人間であると特定している。機械は財産権も、伝統的な人間の寿命も、家族も、居所も、国籍も、意思も、署名も持たない。それに対して、著作権法が著作者が所有すると扱うあらゆる属性を人間が持つ事から、著作権法が人間の著作者性を求めていると読む事は法規の文章、構造及び設計に合致する。人間の著作者性の要件は、要するに、著作権法における共通の語に異例のずれた意味を与えて文章的に木に竹を接ぐ必要性を排除しているのである。アメリカ食品医薬品局対ブラウン&ウィリアムソンタバコ社事件2000年3月21日アメリカ最高裁判決参照(「『法規の語がその文脈において及び全体的な法規の枠組みにおけるその位置に照らして読まればければならないという事は法規解釈の基本的な規範』である。」)(デーヴィス対ミシガン州財務局事件1989年3月28日アメリカ最高裁判決を引用)。

明確化のため、私たちは、これらの法規定のそれぞれが、ある者が著作権の保護を受ける著作物の著者であるために必要な条件を記載しているという事を言っているのではない。著作者には子供も、居所も、伝統的な署名も必要ない。著作権の著作者性のため財産権を所有する能力すら常に求められていた訳ではない。19世紀における妻たちは、カヴァーチャー法により著作権を所有する事を禁じられていたにせよ、著作権保護の対象となる著作物を著作していた。メリッサ・ホームステッド、アメリカの女性著作者及び文学財産権、1822-1869年、第21-62頁;ベルフォード、クラーク社対スクリブナー事件1892年4月11日アメリカ最高裁判決(妻として財産権を所有できなかったにせよ、その本の著作権が侵害された時にターヒューン婦人の著作者性を認めている)参照。

それはともかく、要点は、現在の著作権法の文章は、全体として見たとき、人間性が著作権法における著作物性の必要な条件であると読むのが最も良いという事である。これは「法律全体の規定」が示す読み方である。ジョン・ハンコック相互生命保険社対ハリス信用銀行事件1993年12月13日アメリカ最高裁判決(パイロット生命保険会社対デドー事件1987年4月6日アメリカ最高裁判決を引用)。

 上で訳出した部分に書かれている通り、控訴裁は、アメリカ著作権法の条文を詳細に検討し、機械は権利能力を持たず、著作権保護の前提となる寿命、相続、署名、住所、意図などが機械にはあり得ない事から、著作権法における著作者は人間でしかあり得ない事を明確に導き出している。

 上の部分だけでも十分かも知れないが、アメリカの控訴裁は、AIに著作権を与えない事は創作のインセンティブを削ぐ事になるとのターラー博士の主張についても、以下の通り、きっぱり否定している。

Dr. Thaler also argues that the human-authorship requirement wrongly prevents copyright law from protecting works made with artificial intelligence. Thaler Opening Br. 38.

But the Supreme Court has long held that copyright law is intended to benefit the public, not authors. Copyright law "makes reward to the owner a secondary consideration. *** '[T]he primary object in conferring the monopoly lie[s] in the general benefits derived by the public from the labors of authors.'" United States v. Loew's, Inc., 371 U.S. 38, 46-47 (1962) (quoting Fox Film Co. v. Doyal, 286 U.S. 123, 127 (1932)).

To that public-benefit end, "the law of copyright has developed in response to significant changes in technology." Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417, 430 (1984). Photography, sound recordings, video recordings, and computer programs are all technologies that were once novel, but which copyright law now protects. See Burrow-Giles, 111 U.S. at 58; Goldstein v. California, 412 U.S. 546, 565-566 (1973); Sony, 464 U.S. at 442; Google, 593 U.S. at 21. Importantly, that evolution in copyright protection has been at Congress's direction, not through courts giving new meaning to settled statutory terms.

Contrary to Dr. Thaler's assumption, adhering to the human-authorship requirement does not impede the protection of works made with artificial intelligence. Thaler Opening Br. 38-39.

First, the human authorship requirement does not prohibit copyrighting work that was made by or with the assistance of artificial intelligence. The rule requires only that the author of that work be a human being - the person who created, operated, or used artificial intelligence - and not the machine itself. The Copyright Office, in fact, has allowed the registration of works made by human authors who use artificial intelligence. See Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence, 88 Fed. Reg. 16,190, 16,192 (March 16, 2023) (Whether a work made with artificial intelligence is registerable depends "on the circumstances, particularly how the AI tool operates and how it was used to create the final work.").

To be sure, the Copyright Office has rejected some copyright applications based on the human-authorship requirement even when a human being is listed as the author. See Copyright Office, Re: Zarya of the Dawn (Registration # VAu001480196) (Feb. 21, 2023), https://perma.cc/AD86-WGPM (denying copyright registration for a comic book's images made with generative artificial intelligence). Some have disagreed with these decisions. See Motion Picture Association, Comment Letter on Artificial Intelligence and Copyright at 5 (Oct. 30, 2023), https://perma.cc/9W9X-3EZE (This "very broad definition of 'generative AI' has the potential to sweep in technologies that are not new and that members use to assist creators in making motion pictures."); 2 W. PATRY, COPYRIGHT § 3:60.52 (2024); Legal Professors Amicus Br. 36-37 ("The U.S. Copyright Office guidelines are somewhat paradoxical: human contributions must be demonstrated within the creative works generated by AI.").

Those line-drawing disagreements over how much artificial intelligence contributed to a particular human author's work are neither here nor there in this case. That is because Dr. Thaler listed the Creativity Machine as the sole author of the work before us, and it is undeniably a machine, not a human being. Dr. Thaler, in other words, argues only for the copyrightability of a work authored exclusively by artificial intelligence. Contrast Rearden LLC v. Walt Disney Co., 293 F. Supp. 3d 963 (N.D. Cal. 2018) (holding that companies may copyright work made with motion capture software).

Second, Dr. Thaler has not explained how a ban on machines being authors would result in less original work because machines, including the Creativity Machine, do not respond to economic incentives.

Dr. Thaler worries that the human-authorship requirement will disincentivize creativity by the creators and operators of artificial intelligence. Thaler Opening Br. 36. That argument overlooks that the requirement still incentivizes humans like Dr. Thaler to create and to pursue exclusive rights to works that they make with the assistance of artificial intelligence.

Of course, the Creativity Machine does not represent the limits of human technical ingenuity when it comes to artificial intelligence. Humans at some point might produce creative non-humans capable of responding to economic incentives. Science fiction is replete with examples of creative machines that far exceed the capacities of current generative artificial intelligence. For example, Star Trek's Data might be worse than ChatGPT at writing poetry, but Data's intelligence is comparable to that of a human being. See Star Trek: The Next Generation: Schism (Paramount television broadcast Oct. 19, 1992) ("Felis catus is your taxonomic nomenclature, an endothermic quadruped, carnivorous by nature"). There will be time enough for Congress and the Copyright Office to tackle those issues when they arise.

Third, Congress's choice not to amend the law since 1976 to allow artificial-intelligence authorship "might well be taken to be an acquiescence in the judicial construction given to the copyright laws." White-Smith Music Pub. Co. v. Apollo Co., 209 U.S. 1, 14 (1908). The human-authorship requirement is not new and has been the subject of multiple judicial decisions. The Seventh Circuit has squarely held that authors "of copyrightable works must be human." Kelley v. Chicago Park Dist., 635 F.3d 290, 304 (7th Cir. 2011). And the Ninth Circuit has strongly implied the same when deciding that an author must be a "worldly entity," Urantia Foundation v. Maaherra, 114 F.3d 955, 958 (9th Cir. 1997), and cannot be an animal, Naruto v. Slater, 888 F.3d 418, 426 (9th Cir. 2018).

Finally, even if the human authorship requirement were at some point to stymy the creation of original work, that would be a policy argument for Congress to address. U.S. CONST. Art. I, § 8, cl. 8. "Congress has the constitutional authority and the institutional ability to accommodate fully the varied permutations of competing interests that are inevitably implicated by such new technology." Sony, 464 U.S. at 431.

ターラー博士はまた、人間の著作者性の要件が、著作権法が人口知能によって作られた作品を保護する事を妨げるのは間違っていると主張する。ターラー博士公開書簡。

しかし、アメリカ最高裁は、著作権法は、著作者ではなく、公衆の利益を目的としていると長く判示して着ている。著作権法が「その所有者に報酬を与えるのは二次的考慮である。・・・『独占を付与する事の第一の目的は一般的に著作者の労働から公衆によって導き出される利益にある。』」アメリカ合衆国対ロウズ社事件1962年11月5日アメリカ最高裁判決判決(フォックス映画社対ドワイヤル事件1932年5月16日アメリカ最高裁判決を引用)。

公衆の利益の目標について、「著作権法は技術における大きな変化に対応して発展して来た。」ソニー・アメリカ社対ユニヴァーサル・シティ・スタジオ社事件1984年1月17日アメリカ最高裁判決。写真、録音、録画、そして、コンピュータ・プログラムは全て、かつて新しかったが、今では著作権法により守られている技術である。バロー・ジルス・リソグラフィック社対サロニー事件1884年3月17日アメリカ最高裁判決;ゴールドシュタイン対カリフォルニア事件1973年6月18日アメリカ最高裁判決;上記ソニー・アメリカ社事件アメリカ最高裁判決;グーグル社対オラクル・アメリカ社2021年4月5日アメリカ最高裁判決参照。重要な事として、この著作権保護における発展はアメリカ議会の方針によって来たのであって、裁判所が決まった制定法の用語に新たな意味を与えた事によるのではない。

ターラー博士の推定に反し、人間の著作者性の要件にこだわる事は人工知能とともに作られた作品の保護を妨げるものではない。ターラー博士公開書簡。

第一に、人間の著作者性の要件は、人工知能の補助によるかそれとともに作られた作品に著作権を与える事を禁止していない。この規則は、その作品の著作者が人間-人工知能を作り上げたか、運用したか、利用した者-であって、機械それ自体ではない事を求めているだけである。アメリカ著作権局は、実際、人工知能を利用する人間の著作者により作られた作品の登録を認めて来た。著作権登録指針:人工知能により生成されたマテリアルを含む作品(2023年3月16日)(人工知能とともに作られた作品が登録可能かどうかは「その状況、特にどの様にAIツールが動作するか及びどの様にそれが最終的な作品作り出すのに使われたかによる。」)参照。

確かに、アメリカ著作権局は、人間が著作者としてあげられている場合であっても、人間の著作者性の要件に基づき幾つかの著作権登録申請を拒絶している。アメリカ著作権局、黎明のザリヤに対する決定(2023年10月30日)(生成人工知能によって作られた漫画の画像尾の著作権登録を否定)参照。これらの決定に賛同していない者もいる。モーション・ピクチャー・アソシエーション、人工知能と著作権に関する意見書(2023年10月30日)(この「『生成AI』の非常に広い定義が、構成員が動画を作る際に創作者を支援するのに用いている、新規ではない技術に入り込んで来る可能性がある。」)W.パトリー、著作権法;法学教授アミカス・ブリーフ(「次の様な米著作権局の基準は矛盾しているのではないか:人間の寄与がAIによって作成された創作的作品において示されなければならない。」)参照。

どれほど人工知能が個別の人間の著作者の作品に寄与したかに関するこの線引きの不一致は本事件において何ら関係しない。これはターラー博士が創作機を私たちの前にある作品の唯一の著作者として挙げたからである、それは否定の余地なく機械であり、人間ではない。ターラー博士は、言い換えると、人工知能のみによって著作された作品の著作権保護だけを求めている。コントラスト・レアルデン社対ウォルト・ディズニー社事件2018年2月21日北カリフォルニア地方裁判決(会社は動画ソフトによって作られた作品について著作権を持ち得ると判示)。

第二に、ターラー博士はどうして著作者からの機械の排除が独創的な作品の減少を招くのかについて説明していない、創作機を含む機械は経済的なインセンティブに応えられないのである。

ターラー博士は、人間の著作者性の要件が創作者及び人口知能の運営者による創作のインセンティブを失わさせるだろう事を懸念している。ターラー博士の手紙。この主張は、この要件が、ターラー博士の様な人間に、人工知能の補助とともに作品を作り出し、それにたいする排他的権利を追求するインセンティブをなお与えている事を見逃している。

もちろん、創作機は人工知能に関する場合の人間の技術的な創造力の限界を表すものではない。人間はある点で経済的インセンティブに応える事のできる創作的な非人間を作り出す事ができるであろう。サイエンス・フィクションには現在の生成人工知能の能力を遥かに超える創作機械の例が豊富にある。例えば、スター・トレックのデータは詩を書く事においてChatGPTより悪いであろうが、データの知能は人間のそれに比肩し得る。スター・トレック:次世代:分裂(1992年10月19日放送)(「フェリス・カトゥスは君の分類学的命名であり、内温性の四足獣であり、性質として肉食だ。」)参照。この様な問題が生じる場合、議会と著作権局にはこれに取り組む十分な時間があるだろう。

第三に、1976年以来人工知能の著者を認める法改正をしていないという議会の選択は「著作権法に与えられる司法解釈において暗黙に従うべきものと考え得るだろう。」ホワイト・スミス音楽出版社対アポロ社事件アメリカ最高裁判決1908年2月24日判決。人間の著作者性の要件は新規なものではなく、多くの司法判決の対象となって来た。第7巡回控訴裁は真っ向から「著作権の保護を受ける作品の著作者は人間である」と判示している。ケリー対シカゴ公園区事件2011年2月15日第7巡回控訴裁判決。そして、第9巡回控訴裁は、著作者は「世の中の存在」でなくてはならないと、ウランティア財団対マーヘラ事件1997年6月10日第9巡回控訴裁判決、動物ではあり得ないと、ナルト対スレーター事件2018年4月23日第9巡回控訴裁判決、決定した時、同じ事を強く示唆している。

最後に、仮に人間の著作者性の要件がある点で独自の作品の創作を邪魔し得たとしても、こえは議会が取り組むべき政策論だろう。アメリカ憲法第1条第8節第8項。「議会はその様な新技術によって必然的にもたらされる競合する利害の様々な変更を完全に調整する憲法上の権限及び制度的能力を有する。」ソニー・アメリカ社対ユニヴァーサル・シティ・スタジオ社事件1984年1月17日アメリカ最高裁判決。

 このブログで今まで書いてきた事のほぼ繰り返しになるが、ここで、アメリカの控訴裁が、過去のアメリカ最高裁判決を引用しながら、著作権法による保護の究極的な目的は、あくまで公衆・社会全体の利益であって、著作権者個人の報酬の最大化ではない事を今一度確認している点は重要だろう。

 これは、この4月の日本のAI発明に関する知財高裁判決で(第511回参照)、特許法に関するものだが、「特許権は天与の自然権ではなく、『発明を奨励し、もって産業の発達に寄与する』ことを目的とする特許法に基づいて付与されるもの」と言われた事にも通じるものであって、とにかく個々の知財権の保護強化こそが常に正しいかの様な主張が往々にして罷り通ってしまうのが最近の知財政策決定の間違いの元であると私は断言して憚らない。知財権がない方が良いなどという事を言うつもりはさらさらないが、あらゆる知財権は最終的に公衆の利益に合致するからこそ存在意義があるのであって、そうでなければその権利の強さは公衆の利益を最大化する様に調整されて然るべきものなのである。

 また、これも繰り返しになるので、くどくどと書く事はしないが、このアメリカ控訴裁の判決も、果たして機械に著作権者としての保護を与える事がどうして創作のインセンティブにつながるのか(逆にそうでないとどうして創作のインセンティブを削ぐ事になるのか)について深い疑義を呈した上で、法改正に関する最終的な政策判断は議会によってなされなくてはならないというごく当たり前の事を述べている。

 第508回で取り上げた1月のアメリカ著作権局の報告書には間に合っていなかったものだが、このアメリカ控訴裁の判決でも、AIは著作者たり得ないという事が再確認され、何か判断が分かれたという事はなく、最高裁に上告がされていたとしても、取り上げられて判断が覆る事もないだろう。人間による知的創作の促進を前提とする特許法や著作権法において、AIの様な機械が発明者や著作者にはなり得ないという大原則は、どこの国でも確定していると見て良い。(なお、ターラー氏はアメリカ以外の国では著作権に関する訴訟を起こしていない様だが、基本的に著作権は無方式主義であって、アメリカの様に著作権保護のために著作権局への登録が前提となる様な国がないせいではないかと思う。)

 今後も、特許法や著作権法による保護との関係でAIを利用した場合の発明や創作について世界的に議論が続けられるのではないかと思うが、どこの国の司法判断でも創作保護法としての知的財産法の原則が確認されている中、日本でもその原則を踏まえた丁寧な政策検討がなされる事を願っている。

| | コメント (0)

2025年5月18日 (日)

第512回:人口知能(AI)の学習のための著作物の利用に関するアメリカ著作権局のプレ公開版の報告書第3部

 もう少し時間が掛かると思っていたが、先週、アメリカ著作権局は、第508回で取り上げた著作権保護適格性に関する第2部に続き、人口知能(AI)の学習のための著作物の利用に関する報告書第3部をプレ公開版としてそのAI関連ページで公表した。

 このアメリカ著作権局のプレ公開版報告書第3部(pdf)は、その結論に以下の様に書かれている通り、AI学習のための著作物の利用が著作権侵害となるかどうかはそれがフェアユースと評価されるかどうかにに懸かっている事、今までの判例や意見などをまとめながら、フェアユースとなる場合もならない場合もあるだろうという事を言うに留まり、当然と言えば当然だが、この利用形態について決定的な判断を示しているなどという事はない。(以下、いつも通り翻訳は全て拙訳。)

VI. CONCLUSION

Throughout its history, copyright law has adapted to new technology, furthering its progress while preserving incentives for creative activity. This has enabled our nation's creative and technology industries to become global leaders in their fields. While the use of copyrighted works to power current generative AI systems may be unprecedented in scope and scale, the existing legal framework can address it as in prior technological revolutions. The fair use doctrine in particular has served to flexibly accommodate such change. We believe it can do so here as well.

In applying current law, we conclude that several stages in the development of generative AI involve using copyrighted works in ways that implicate the owners' exclusive rights. The key question, as most commenters agreed, is whether those acts of prima facie infringement can be excused as fair use.

The fair use determination requires balancing multiple statutory factors in light of all relevant circumstances. Although it is not possible to prejudge the result in any particular case, precedent supports the following general observations:

Various uses of copyrighted works in AI training are likely to be transformative. The extent to which they are fair, however, will depend on what works were used, from what source, for what purpose, and with what controls on the outputs - all of which can affect the market. When a model is deployed for purposes such as analysis or research - the types of uses that are critical to international competitiveness - the outputs are unlikely to substitute for expressive works used in training. But making commercial use of vast troves of copyrighted works to produce expressive content that competes with them in existing markets, especially where this is accomplished through illegal access, goes beyond established fair use boundaries.

For those uses that may not qualify as fair, practical solutions are critical to support ongoing innovation. Licensing agreements for AI training, both individual and collective, are fast emerging in certain sectors, although their availability so far is inconsistent. Given the robust growth of voluntary licensing, as well as the lack of stakeholder support for any statutory change, the Office believes government intervention would be premature at this time. Rather, licensing markets should continue to develop, extending early successes into more contexts as soon as possible. In those areas where remaining gaps are unlikely to be filled, alternative approaches such as extended collective licensing should be considered to address any market failure.

In our view, American leadership in the AI space would best be furthered by supporting both of these world-class industries that contribute so much to our economic and cultural advancement. Effective licensing options can ensure that innovation continues to advance without undermining intellectual property rights. These groundbreaking technologies should benefit both the innovators who design them and the creators whose content fuels them, as well as the general public.

Finally, as in prior Parts of this Report, the Office recognizes that facts on the ground are evolving at a rapid pace. We will continue to monitor developments in technology, case law, and markets, and to offer further assistance to Congress as it considers these issues.

Ⅵ.結論

その歴史を通じて、著作権は新技術に適応して来ており、さらに創作活動のためのインセンティブを保持しながらその歩みを進めている。これは私たちの国の創作的技術産業がその分野において世界的リーダーとなる事を可能として来た。現在の生成AIシステムを作動させるための著作物の利用は、範囲と規模において前例のないものであろうが、現行の法的枠組みは過去の技術革新と同じくそれに対処できる。フェアユースの法理は特にこの様な変化に柔軟に対応するために役に立って来た。私たちは今回もそれは同じ様にできると信じる。

現行法を適用するに際し、私たちは生成AIの開発における幾つかの段階に権利者の排他的権利に関わるやり方で著作物を利用する事が含まれていると結論する。主たる問題は、ほとんどの意見提出者が認める通り、この一応の侵害行為がフェアユースとして許され得るかである。

フェアユースの判断には関係するあらゆる状況に照らして法定の複数の要素を衡量する事が求められる。特定のケースにおける結果を予断する事はできないが、先例は次の一般的な所見を支持する:

AI学習における著作物の様々な利用は変形的なものとなる事が多いだろう。それがどれほど公正であるかは、しかしながら、どの様な元から、どの様な目的で、どの様な作品が使われているのか、そして、その出力にどの様な制御があるのか-市場に影響し得るあらゆる事-に懸かっている。モデルが分析や研究-国際競争において重要なものである利用の形-の様な目的のために展開される場合、出力は学習に用いられた表現的作品を置き換える事はありそうにない。しかし、既存の市場において競合する表現的コンテンツを製造するために著作物の大宝庫を商業利用する事は、特にこれが違法アクセスを通じて達成される場合、確立されたフェアユースの境界を超える。

公正とは評価され得ないこれらの利用について、進展するイノベーションにとって実務的な解決が重要である。その利用可能性はまちまちであるが、AI学習のための個別及び包括の両方のライセンス契約が幾つかの分野において急速に立ち上がっている。自主的なライセンスが安定的に成長している事とともに、法令の変更に対する関係者の支持が欠けている事からも、アメリカ著作権局は政府の介入は現時点で時期尚早であると考える。むしろ、ライセンス市場こそ、できる限り早く初期の成功をさらなる文脈に広げ、発展を続けるべきである。残る隔たりが埋められそうにない領域において、拡大集中ラリセンスの様な代替アプローチが市場の失敗に対応するために検討されるべきである。

私たちの見解では、AI領域におけるアメリカのリーダーシップは、私たちの経済的及び文化的進展に大いに寄与している、これらの世界レベルの産業の両方によって支持され、進められるのが最良であろう。効果的なライセンスの選択肢はイノベーションが知的財産権を危うくする事なく進展し続ける事を保証する。この画期的な技術はそれを設計するイノベーター及びそのコンテンツがそれを加速するクリエーター並びに一般公衆をともに裨益するべきものである。

最後に、本報告の先行部分同様、アメリカ著作権局はこの領域における事実は急速に進展している事を認める。私たちは技術における発展、判例及び市場を注視し続け、議会がこの問題を検討するに際しさらなる支援を提供し続けるであろう。

 同時にアメリカ著作権局長のパールムッター氏が解任されたという事も報道されており、この報告書の内容が権利者寄りである事がトランプ現米大統領とその幹部の癇に障ったので解任されたなどという事も噂されているのだが、上でも書いた通り、この報告書の内容は今までの判例などから中立的に言える事を言ったに留まり、特に権利者寄りであるとも、AI事業者・利用者寄りであるとも言い難い。トランプ政権の主要幹部が1日でこの大部の報告書の内容をまともに評価して著作権局長の解任を決めたなどという事はありそうになく、おそらく何かの事情で近く解任される事を知ったパールムッター氏が第3部を急ぎプレ公開版として公表したという方が妥当なのではないかと私は思っている。

 上の結論の部分で、「既存の市場において競合する表現的コンテンツを製造するために著作物の大宝庫を商業利用する事は、特にこれが違法アクセスを通じて達成される場合、確立されたフェアユースの境界を超える」と書かれているのも、今までのフェアユースに関する判例などを踏まえて考えればほぼ当たり前の事に近く、同一市場で競合する類似コンテンツを生成するため、違法アクセスにより入手した大量の著作物をそうと知ってAI学習に利用する事がフェアユースに該当するなどと心の底から思っている法律の専門家はアメリカでもいないに違いない(置かれた立場からそう主張する事はあるかも知れないが)。

 最終的にはアメリカ最高裁により決められる事になるであろうが、この報告書でまとめられていることを見ても、AI学習への著作物利用のフェアユース該当性に関する判断は、既存の著作物に依拠して類似した物を意図的に出力させようとするものであるかどうか、著作権侵害とならない出力を作り出そうとするものであるかどうかが大きな判断要素となり得るだろうと思え、日本の著作権法第30条の4の非享受目的利用の例外や、ヨーロッパにおけるテキスト及びデータマイニングの例外の適用の判断とも収斂して行くのではないかと私は期待している。(日欧でもまだ判例はほぼ出ていない状況だが、それぞれ、日本の著作権法における考え方をまとめた文化庁の報告書について第492回、欧州新著作権指令との関係について第481回参照。なお、まだ地裁レベルだが、第504回で取り上げた様に、ドイツにおいて非営利でのAI学習用データの提供は研究目的データマイニングの権利制限に該当するとした判決が存在している。)

 また、このアメリカ著作権局の報告書は全て読むに値するものと思うが、フェアユースの4つの法定要素、(1)商業的かどうかなど利用の目的及び性格、(2)著作物の性質、(3)著作物全体に対する利用部分の量と本質性、(4)著作物の市場又は価値に対する利用の潜在的影響の中でも、ここでは、特に評価が微妙な最初の要素における変形性と違法アクセスについて述べた部分について以下に訳出しておく。

IV. FAIR USE

...

A. Factor One

...

2. Transformativeness

...

As discussed above, Warhol requires examining not just the immediate act of copying but its ultimate goal.[Note 259: See Warhol, 598 U.S. at 550-51; see also Authors Guild, Inc. v. HathiTrust, 755 F.3d 87, 97-105 (2d Cir. 2014) (looking beyond copying books for "digitization" and "storage" to consider the ultimate purposes of those uses: full-text search functionality, accessibility for the print-disabled, and preservation); Sony Comput. Ent., Inc. v. Connectix Corp., 203 F.3d 596, 601, 606-07 (9th Cir. 2000) (evaluating purpose of copying to reverse engineer software by considering the final product that the reverse engineering enabled); Infinity Broad. Corp. v. Kirkwood, 150 F.3d 104, 108-09 (2d Cir. 1998); A&M Recs., Inc. v. Napster, Inc., 239 F.3d 1004, 1015 (9th Cir. 2001); Flo & Eddie, Inc. v. Sirius XM Radio, Inc., 821 F.3d 265, 270 n.4 (2d Cir. 2016) ("The fair-use analysis applicable to [creating an internal database of recordings] … is bound up with whether the ultimate use of the internal copies [to make public performances] is permissible."). See supra Section IV.A.1.] Accordingly, whether copying a work to compile a training dataset is transformative depends on whether the dataset will be used for a transformative purpose.

In the Office's view, training a generative AI foundation model on a large and diverse dataset will often be transformative. The process converts a massive collection of training examples into a statistical model that can generate a wide range of outputs across a diverse array of new situations. It is hard to compare individual works in the training data - for example, copies of The Big Sleep in various languages - with a resulting language model capable of translating emails, correcting grammar, or answering natural language questions about 20th-century literature, without perceiving a transformation. The purpose of creating works of authorship is to disseminate them for human enjoyment and education. Many AI models, however, are meant to perform a variety of functions, some of which may be distinct from the purpose of the copyrighted works they are trained on.[Note 260: Cf. Authors Guild v. Google, Inc. (Google Books), 804 F.3d 202, 216-18 (2d Cir. 2015); HathiTrust, 755 F.3d at 97; Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d at 1165; Kelly v. Arriba Soft Corp., 336 F.3d 811, 819 (9th Cir. 2003).] For example, a language model can be used to help learn a foreign language by chatting with users on diverse topics and offering corrective feedback.[Note 261: Cf. Stavros Athanassopoulos, The use of ChatGPT as a learning tool to improve foreign language writing in a multilingual and multicultural classroom, Advances in Mobil Learning Educational Research (2023), https://www.syncsci.com/journal/AMLER/article/view/AMLER.2023.02.009; see also OpenAI and the California State University System Bring AI to 500,000 Students and Faculty, OPENAI (Feb. 4, 2025), https://openai.com/index/openaiand-the-csu-system/; Linda Kane, Space Force Generative AI Challenge Empowers Guardians Through Education, Collaboration, U.S. SPACE FORCE (Dec. 10, 2024), https://www.spaceforce.mil/News/Article-Display/Article/3992437/space-force-generative-ai-challenge-empowers-guardians-through-education-collab/; Severin Rodler et al., Generative Artificial Intelligence in Surgery, 175 SURGERY 1496 (2024), https://www.sciencedirect.com/science/article/abs/pii/S0039606024001193?via%3Dihub.]

But transformativeness is a matter of degree, and how transformative or justified a use is will depend on the functionality of the model and how it is deployed. On one end of the spectrum, training a model is most transformative when the purpose is to deploy it for research,[Note 262: Cf. Google Books, 804 F.3d at 217 ("[T]he purpose of Google's copying of the original copyrighted books is to make available significant information about those books, permitting a searcher to identify those that contain a word or term of interest, as well as those that do not include reference to it. In addition, through the ngrams tool, Google allows readers to learn the frequency of usage of selected words in the aggregate corpus of published books in different historical periods. We have no doubt that the purpose of this copying is the sort of transformative purpose described in Campbell as strongly favoring satisfaction of the first factor.").] or in a closed system that constrains it to a non-substitutive task. For example, training a language model on a large collection of data, including social media posts, articles, and books, for deployment in systems used for content moderation does not have the same educational purpose as those papers and books.

On the other end of the spectrum is training a model to generate outputs that are substantially similar to copyrighted works in the dataset. For example, a foundation image model might be further trained on images from a popular animated series and deployed to generate images of characters from that series. Unlike cases where copying computer programs to access their functional elements was necessary to create new, interoperable works, using images or sound recordings to train a model that generates similar expressive outputs does not merely remove a technical barrier to productive competition. In such cases, unless the original work itself is being targeted for comment or parody, it is hard to see the use as transformative.[Note 263: The decision to train on expressive works when there are available alternatives may itself reflect a lack of transformative purpose. For example, an image model could be trained on mass image data collected through automated means (street-view cars, body cameras, security cameras), yet developers often choose aesthetic images such as stock photography. This suggests the purpose is not simply to generate images of the physical world, but to generate images that have expressive qualities like the originals.]

Many uses fall somewhere in between. The use of a model may share the purpose and character of the underlying copyrighted works without producing substantially similar content. Where a model is trained on specific types of works in order to produce content that shares the purpose of appealing to a particular audience, that use is, at best, modestly transformative. Training an audio model on sound recordings for deployment in a system to generate new sound recordings aims to occupy the same space in the market for music and satisfy the same consumer desire for entertainment and enjoyment. In contrast, such a model could be deployed for the more transformative purpose of removing unwanted distortion from sound recordings.

Because generative AI models may simultaneously serve transformative and nontransformative purposes,[Note 264: As described above, the strength of generative pre-trained language models is their ability to perform well on a variety of tasks when given natural language directions. See supra notes 39-41, 102-103.] restrictions on their outputs can shape the assessment of the purpose and character of the use. As described above, developers can apply training techniques or deployment guardrails so that the model rejects requests for excerpts of copyrighted works or even refuses to generate expressive works. Where such restrictions are effective, the system will be less capable of fulfilling the purpose of the original works, and their use in training may be more transformative.

...

4. Unlawful Access

...

In the Office's view, the knowing use of a dataset that consists of pirated or illegally accessed works should weigh against fair use without being determinative.[Note 292: See Harper & Row Publishers, Inc. v. Nation Enters., 471 U.S. 539, 562-63 (1985) (finding a user's knowing exploitation of a "purloined manuscript" to be relevant to the character of the use); Chi. Bd. of Educ. v. Substance, Inc., 354 F.3d 624, 628 (7th Cir. 2003) (explaining that the fact that the access was unauthorized "does not exclude the possibility of a fair use defense."); L.A. News Serv. v. KCAL-TV Channel 9, 108 F.3d 1119, 1122 (9th Cir. 1997) (finding relevant to the propriety of the user's conduct that, after being refused a license, defendant then "obtained a copy of the tape ... directly copied the original, superimposed its logo ... and used it for the same purpose for which it would have been used had it been paid for"); Atari Games Corp. v. Nintendo of Am. Inc., 975 F.2d 832, 843 (Fed. Cir. 1992) ("To invoke the fair use exception, an individual must possess an authorized copy of a literary work."); cf. Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d at 1164 n.8 ("[W]e conclude that Google's inclusion of thumbnail images derived from infringing websites in its Internet-wide search engine activities does not preclude Google from raising a fair use defense.").] Courts have expressed some uncertainty about whether good or bad faith generally is relevant to the fair use analysis.[Note 293: See Google LLC v. Oracle Am., Inc., 593 U.S. at 32-33 ("As for bad faith, our decision in Campbell expressed some skepticism about whether bad faith has any role in a fair use analysis. We find this skepticism justifiable, as '[c]opyright is not a privilege reserved for the well-behaved.'" (citations omitted)); Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 585 n.18 (1994) ("[W)e reject [plaintiff's] argument that [defendant's] request for permission to use the original should be weighed against a finding of fair use."); see also Pierre N. Leval, Campbell As Fair Use Blueprint?, 90 WASH. L. REV. 597, 612-14 (2015); Simon Frankell & Matt Kellogg, Bad Faith and Fair Use, 60 J. COPYRIGHT SOC'Y USA 1 (2013), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2165468; Pierre N. Leval, Toward A Fair Use Standard, 103 HARV. L. REV. 1105, 1126-28 (1990).] The cases in which they have done so, however, involved defendants who used copyrighted works despite the owners' denial of permission. Training on pirated or illegally accessed material goes a step further.[Note 294: In addition, some commenters suggested that, to the extent copyright owners "opt out" of having their material used to train AI, whether through terms of use, the robots.txt instructions, or other means, a defendant's decision to ignore such opt-outs might inform the fair use analysis. See, e.g., N/MA Initial Comments at 44; Pamela Samuelson et al. Initial Comments at 24 ("A defendant's ... disregard of robots.txt exclusions and similar mechanisms could each be framed in terms of an argument against fair use under the fourth factor.").] Copyright owners have a right to control access to their works, even if someone seeks to obtain them in order to make a fair use.[Note 295: Cf. 17 U.S.C. § 1201 (prohibiting circumvention of technological protection measures used by copyright owners to control access to their works); Samuelson, et al. Initial Comments at 24 ("One might argue that although copyright owners do not have a right to charge for fair uses as such, they do have a right to charge for access to their works. As such, it may be deemed harmful or unfair for commercial users to bypass the market for access to train their LLMs without a compelling reason.").] Gaining unlawful access therefore bears on the character of the use.

Ⅳ.フェアユース

(略)

A.要素1

(略)

2.変形性

(略)

上で議論した通り、ウォーホル事件最高裁判決は単に直近の複製行為ではなくその究極的な目標を見る事を求めている。[原注259:2023年5月18日のアンディ・ウォーホル財団対ゴールドスミス事件アメリカ最高裁判決;以下も参照、2014年6月10日の全米作家協会対ハイチ・トラスト事件第2巡回控訴裁判決(「デジタル化」及び「蓄積」のための本の複製行為を超えてその利用の究極的な目的である、全文サーチ、視覚障害者のためのアクセス性、保存を考慮);2000年2月10日のソニー・コンピュータ・エンタテイメント社対コネティックス社事件第9巡回控訴裁判決(リバースエンジニアリングが可能とした最終製品を考慮し、ソフトウェアをリバースエンジニアするための複製行為の目的を評価);2001年2月12日のA&Mレコード社対ナップスター社第9巡回控訴裁判決;2016年12月20日のフロ&エディー社対シリウスXMラジオ社事件第2巡回控訴裁判決(「[録音の内部データベースを作成する事]に対して適用されるフェアユース分析は…[公演をするという]その内部コピーの究極的な利用が許されるかどうかに密接に関係する。」)。上記Ⅳ.A.1節も参照。]したがって、学習データセットをまとめるために作品を複製する事が変形的であるかはそのデータセットが変形的な目的で利用されるかに懸かっている。

アメリカ著作権局の見解では、大規模で多様なデータセットによる生成AIの基本モデルの学習は変形的である事が多いであろう。そのプロセスは学習例の大量のコレクションを新しい状況の多様な幅に渡り広い範囲で生成できる統計モデルへと変換する。その学習データにおける個々の作品-例えば、様々な言語における大いなる眠りの複製-を、変形を認める事なく、電子メールの翻訳、文法の訂正又は20世紀文学に関する自然言語の質問に対する回答と比較する事は困難である。著作者性のある作品を作り出す目的は人間の享受と教育のためにそれを広める事にある。多くのAIモデルは、しかしながら、様々な機能を実行する事を目指しており、それらの内の幾つかは学習される著作物の目的とは区別され得るであろう。[原注260:2015年10月16日の全米作家協会対グーグル社(グーグルブックス)事件第2巡回控訴裁判決;ハイチ・トラスト事件第2巡回控訴裁判決;2007年5月16日のパーフェクト10社対アマゾン・ドットコム社事件第9巡回控訴裁判決;2003年7月7日のケリー対アリバ・ソフト社事件第9巡回控訴裁判決。]例えば、言語モデルは、様々なトピックについて利用者とチャットし、訂正フィードバックを提供する事により、外国語の学習を助ける事に使われ得る。[原注261:(訳注:参考文献は上の原文参照。)]

しかし、変形性は程度の問題であり、どれほど変形的であるか又は利用が正当化されるかは、モデルの機能性及びどの様にそれが展開されるかによるであろう。このスペクトルの一端において、その目的がそれを研究のために展開するか[原注262:グーグルブックス事件第2巡回控訴裁判決(「元の著作権の保護を受ける本を複製するグーグルの目的は、これらの本に関する有意な情報を入手可能とし、研究者に関心のある語又は用語を含むもの並びにそれに対する参照を含まないものを特定する事を可能とする事にある。さらに、ngramツールによって、グーグルは異なる時代に出版された本の集合コーパスにおける選択された語の利用の頻度を学ぶ事を読者に可能としている。この複製行為の目的はキャンベル事件最高裁判決において述べられた種類の変形目的であり、第1の要素を強く肯定的に満たすものである事に疑いの余地はない。」)]、非代替タスクのためにそれを含むクローズドシステムに展開する事にある場合、モデルの学習はほぼ変形的であろう。例えば、コンテンツモデレーションのために使われるシステムのために展開される、ソーシャルメディアの投稿、記事及び本を含む大規模コレクションによる言語モデルの学習はこれらの論文や本と同じ教育的な目的を持たない。

そのスペクトルの他端には、データセットにおける著作物に実質的に類似する出力を生成するためにモデルを学習させる事がある。例えば、基本画像モデルを、人気のアニメシリーズの画像でさらに学習させ、そのシリーズのキャラクターの画像を生成するために展開する事があり得るだろう。その機能的要素にアクセスするためのコンピュータプログラムの複製行為が、新しく相互運用可能な作品を作り出すために必要な場合と異なり、類似した表現的出力を生成するために画像又は録音を利用する事は、単に生産的競争の技術的障害を取り除く事ではない。この様な場合、元の作品それ自体が批評又はパロディのターゲットとなっていない限り、その利用を変形的と見るのは困難である。[原注263:代替物が入手可能である場合に表現的作品による学習する事を決定する事自体、変形目的の欠如を示すものであり得るだろう。例えば、画像モデルは自動手段によって収集された大量の画像データ(ストリートビュー車、ボディカメラ、セキュリティカメラ)で学習され得るが、開発者はストック写真の様な美的画像を選択する事が良くある。この事はその目的が現実世界の画像を生成する事にあるのではなく、オリジナルの様な表現的性質を持つ画像を生成する事にあるのを示唆する。]

多くの利用は何かしらその間にある。モデルの利用は、実質的に類似するコンテンツを作成しないが、基礎となる著作物の目的と性格を共有し得る。特定の視聴者の気を引く目的で共有されるコンテンツを生成するためにモデルが特定の型の作品により学習される場合、その利用は、せいぜい、多少変形的なものである。新しい録音を生成するためのシステムへの展開のために録音によりオーディオモデルを学習させる事は、音楽市場において同じ場を占め、娯楽と享受のための同じ消費者の望みを満たす事を目的としている。対照的に、その様なモデルは、録音から望ましくない歪みを取り除くというより変形的な目的のために展開され得るであろう。

生成AIモデルは同時に変形的及び非変形的目的に使われ得るものであるから[原注264:上で述べた通り、事前学習言語モデルの強さとは、自然言語の指示が与えられたときに多様な仕事を実行できるその能力である。前注39-41、102-103参照。]、その出力に対する制限はその利用の目的及び性格の評価を形作り得る。上で述べた通り、モデルが著作物の抜粋に対する要求を拒絶し、さらに表現的作品を生成する事を拒絶する様にも、開発者は学習技術や展開の安全策を展開できるのである。この様な制限が有効である場合、システムは元の作品の目的を満たす事がよりできなくなるであろうし、学習におけるその利用はより変形的なものであり得るであろう。

(略)

4.違法アクセス

(略)

アメリカ著作権局の見解では、海賊版又は違法にアクセスされた作品からなるデータセットのそうと知っての利用は、決定的なものではないが、フェアユースに否定的に働くものである。[原注292:1985年5月20日のハーパー&ロー出版社対ナショナル・エンタープライズ事件アメリカ最高裁判決(「盗まれた原稿」のそうと知っての利用はその利用の性格に関係すると認定);2003年12月31日のシカゴ教育委員会対サブスタンス社事件第9巡回控訴裁判決(アクセスが不正のものであったという事実は「フェアユースの抗弁の可能性を排除しない」と説明);1997年3月11日のロサンゼルス・ニュース・サービス対KCAL-TVチャンネル9事件第9巡回控訴裁判決(ライセンスを拒絶した後で、被告が「テープの複製を入手し・・・オリジナルを直接コピーし、そのロゴを入れ・・・支払って使われたであろうのと同じ目的でそれを使った」という利用者の行為の妥当性に関する認定);1982年9月10日のアタリゲームズ社対米ニンテンドー社事件連邦巡回控訴裁判決(「フェアユースの例外を持ち出すためには、個人は文芸作品の正規の複製を所有していなければならない。」);パーフェクト10社対アマゾン・ドットコム社事件第9巡回控訴裁判決(「私たちは、グーグルのインターネットに渡る検索エンジンの活動において侵害ウェブサイトに由来するサムネイルが含まれる事はグーグルがフェアユースの抗弁を提起するのを妨げない。」)。]裁判所は善意又は悪意が一般的にフェアユース分析に関係するかどうかについて多少のばらつきを示して来た[原注293:2021年4月5日のグーグル社対米オラクル社事件アメリカ最高裁判決(「悪意に関して、キャンベル事件における私たちの決定は悪意がフェアユース分析において何らかの役割を果たすかについて懐疑を示している。『著作権は良く振る舞う者に取り置かれる特権ではない』のであり、私たちはこの懐疑は正当化され得るものと考える。」);1994年3月7日のキャンベル対エイカフ=ローズ・ミュージック社事件アメリカ最高裁判決(「私たちはオリジナルを利用する許可の求めがフェアユースの認定に否定的に働くという[被告の]主張に同意しない。」);(訳注:その他参考文献については上の原文参照。)]。しかしながら、裁判所がそうして来た事件は、権利者による許可の拒絶にも関わらず著作物を利用した被告に関するものである。海賊版又は違法にアクセスされたマテリアルによる学習はさらに一歩先を行く[原注294:さらに、幾つかの意見は、著作権者のそのマテリアルがAIの学習のために使われる事に対する「オプトアウト」の程度、利用規約、robot.txtによる指示又は他の手段を通じて、被告がその様なオプトアウトを無視する決定をしているかどうかがフェアユース分析に情報を与え得るであろうと示唆している。例えば、N/MA初期意見、第44ページ;パメラ・サミュエルソン他初期意見、第24ページ(「被告の…robot.txtによる除外及び類似のメカニズムの無視はそれぞれ第4の要素においてフェアユースに否定的な主張となり得るであろう。」)参照。]。誰かが公正に利用するためにそれを入手しようとして来た場合であっても、著作権者はその作品に対するアクセスをコントロールする権利を有する[原注295:アメリカ著作権法第1201条(その作品へのアクセスをコントロールするために著作権者により利用される技術的保護手段の回避の禁止)参照;サミュエルソン他初期意見、第24ページ(「著作権者がフェアユース自体について課金する権利を持たないにせよ、その作品へのアクセスについて課金する権利を有すると主張し得るであろう。そうであるならば、強行的な理由なくその大規模言語モデルを学習させるために商業的利用者がアクセス市場を迂回するのは有害か不公正なものとみなされるべきであろう。」)。]。したがって、違法アクセスの入手はその利用の性格に影響する。

 ここで書かれている事も特におかしな事はなく、今までの様々なアメリカの主要な著作権判例、特に2023年のウォーホル事件アメリカ最高裁判決(ウォーホルによるプリンス作品への写真の利用はフェアユースに十分なほど変形的でないとしたケース、英語版Wikipedia参照)、2021年のグーグル対オラクル事件最高裁判決(JavaのAPI利用はフェアユースであると判断したケース、第440回参照、英語版Wikipediaも参照)、1994年のキャンベル事件最高裁判決(商業的パロディはフェアユースに該当すると判断したケース、英語版Wikipedia参照)などに照らし、AI学習への著作物の利用をどの様に考えるかをあり得るやり方で書いていると思うが、この報告書の分析は第1の要素の中でもウォーホル事件最高裁判決に依拠して変形性を若干強調し過ぎている様に私には見える。これはアメリカにおける最も最近の判例なのでやむを得ないかも知れないが、私自身はウォーホル事件におけるアメリカ最高裁の多数意見の判断は間違っていると思っているし、AI学習自体が新しい利用形態である事、フェアユースの判断においては常に新たな状況に応じて各要素を総合考慮すべきである事を忘れてはならないだろう。(アメリカのウォーホル事件についてはここで取り上げていなかったのだが、ウォーホル事件に対する私の考えとそこからAI学習のための著作物の利用についてどう考えるべきかという事はできればまたどこかで書きたいと思っている。)

 繰り返しになるが、このアメリカ著作権局のプレ公開版報告書第3部は、全体として、今までのアメリカの判例などに照らし、AI学習のための著作物の利用はフェアユースとなる場合もならない場合もあるだろう事を詳細に中立的にまとめた資料として精読に値するものである。ただ、上の変形性に関する部分など、検討の詰めが若干甘い様に見える所はプレ公開版として急ぎ公表した影響もあったのかも知れず、著作権局長の解任によって正式版の公表がもしかしたら遅れるか、公表されないかも知れないのはかなり残念である。

 知財本部の次回5月19日の構想委員会の議事(開催案内参照)を見ると、そろそろ今年の知財計画が決定されるのではないかと思えるので、次回おそらく知財計画の文章の確認をしてから、またAIと知的財産の関係に関する国際動向について書いて行きたいと思っている。

(2025年5月25日の追記:この前見た時には書かれていなかったと思うが、アメリカ著作権局のAI関連ページに、第3部の最終版は分析や結論に実質的な変化のない見込みで近い将来公表されると書かれているので、最終版が公表されないということはなさそうである。)

| | コメント (0)

2025年2月 9日 (日)

第508回:人口知能(AI)を利用した場合の著作権保護適格性に関するアメリカ著作権局の報告書第2部

 先月1月29日に、アメリカ著作権局が、デジタルレプリカ(ディープフェイク)に関する第1部に続いて、人口知能(AI)を利用した場合の著作権保護適格性に関する報告書の第2部を公表した。この報告書は特に新しい事が書かれているものではないが、アメリカにおけるAIと著作権の関係に関する最新の整理を示したものとして一読に値するものであり、ここで取り上げたいと思う。(第1部については第501回参照。)

 この報告書第2部(pdf)(アメリカ著作権局のリリースも参照)で、特にAIを利用した場合の著作権保護適格性、著作物性について書いているのは第Ⅱ章「著作者による著作と人工知能」であり、そこで、著作権の保護は人間の著作に与えられ、人間でないものによって作り出された物に著作権を認めた裁判所はないという極めて基本的な法原則を示した上で、AIシステムの補助的利用、プロンプト、表現的入力、AI生成物の修正やアレンジという4つの主なAI利用のパターンについて以下の様にまとめている。(以下、ポイントとなる部分について抄訳する。)

C. Assistive Uses of AI Systems

...

The Office agrees that there is an important distinction between using AI as a tool to assist in the creation of works and using AI as a stand-in for human creativity. While assistive uses that enhance human expression do not limit copyright protection, uses where an AI system makes expressive choices require further analysis. This distinction depends on how the system is being used, not on its inherent characteristics.

...

D. Prompts

...

2. Analysis

The Office concludes that, given current generally available technology, prompts alone do not provide sufficient human control to make users of an AI system the authors of the output. Prompts essentially function as instructions that convey unprotectible ideas. While highly detailed prompts could contain the user's desired expressive elements, at present they do not control how the AI system processes them in generating the output.

Cases regarding joint authorship support this conclusion. These cases address the amount of control that is necessary to claim authorship. The provision of detailed directions, without influence over how those directions are executed, is insufficient. As the Third Circuit explained, when a person hires someone to execute their expression, "that process must be rote or mechanical transcription that does not require intellectual modification or highly technical enhancement" for the delegating party to claim copyright authorship in the final work.

Although entering prompts into a generative AI system can be seen as similar to providing instructions to an artist commissioned to create a work, there are key differences. In a human-to-human collaboration, the hiring party is able to oversee, direct, and understand the contributions of a commissioned human artist. Depending on the nature of each party's contributions, the artist may be the sole author, or the outcome may be a joint work or work made for hire. In theory, AI systems could someday allow users to exert so much control over how their expression is reflected in an output that the system's contribution would become rote or mechanical. The evidence as to the operation of today's AI systems indicates that this is not currently the case. Prompts do not appear to adequately determine the expressive elements produced, or control how the system translates them into an output.

The gaps between prompts and resulting outputs demonstrate that the user lacks control over the conversion of their ideas into fixed expression, and the system is largely responsible for determining the expressive elements in the output. In other words, prompts may reflect a user's mental conception or idea, but they do not control the way that idea is expressed. This is even clearer in the case of generative AI systems that modify or rewrite prompts internally.

The following image, which the Office generated by entering a prompt into a popular commercially available AI system, illustrates this point:

Prompt

This prompt describes the subject matter of the desired output, the setting for the scene, the style of the image, and placement of the main subject. The resulting image reflects some of these instructions (e.g., a bespectacled cat smoking a pipe), but not others (e.g., a highly detailed wood environment). Where no instructions were provided, the AI system filled in the gaps.

For instance, the prompt does not specify the cat’s breed or coloring, size, pose, any attributes of its facial features or expression, or what clothes, if any, it should wear beneath the robe. Nothing in the prompt indicates that the newspaper should be held by an incongruous human hand.

The fact that identical prompts can generate multiple different outputs further indicates a lack of human control. As one popular system explains on its website, "[n]o matter how detailed ... the same text describes an infinite number of possible" outputs. In these circumstances, the black box of the AI system is providing varying interpretations of the user's directions.

Repeatedly revising prompts does not change this analysis or provide a sufficient basis for claiming copyright in the output. First, the time, expense, or effort involved in creating a work by revising prompts is irrelevant, as copyright protects original authorship, not hard work or "sweat of the brow." Second, inputting a revised prompt does not appear to be materially different in operation from inputting a single prompt. By revising and submitting prompts multiple times, the user is "re-rolling" the dice, causing the system to generate more outputs from which to select, but not altering the degree of control over the process. No matter how many times a prompt is revised and resubmitted, the final output reflects the user's acceptance of the AI system's interpretation, rather than authorship of the expression it contains.

...

E. Expressive Inputs

...

As an example, in the following work submitted to the Office for registration, the author had created a hand-drawn illustration and used it as an input, along with the prompt shown below.

The AI system produced this output:

Expressive

The drawing itself is a copyrightable work, and its expressive elements are clearly perceptible in the output, including the outline of the mask, the position of the nose, mouth, and cheekbones relative to the shape of the mask, the arrangement of the stems and rosebuds, and the shape and placement of the four leaves.

The applicant disclaimed "any non-human expression" appearing in the final work,such as the realistic, three-dimensional representation of the nose, lips, and rosebuds, as well as the lighting and shadows in the background. After reviewing the information provided in the application, the Office registered the work with an annotation stating: "Registration limited to unaltered human pictorial authorship that is clearly perceptible in the deposit and separable from the non-human expression that is excluded from the claim."

As illustrated in this example, where a human inputs their own copyrightable work and that work is perceptible in the output, they will be the author of at least that portion of the output. Their own creative expression will be protected by copyright, with a scope analogous to that in a derivative work. Just as derivative work protection is limited to the material added by the later author, copyright in this type of AI-generated output would cover the perceptible human expression. It may also cover the selection, coordination, and arrangement of the human-authored and AI-generated material, even though it would not extend to the AI-generated elements standing alone.

F. Modifying or Arranging AI-Generated Content

...

Many popular AI platforms offer tools that encourage users to select, edit, and adapt AI-generated content in an iterative fashion. Midjourney, for instance, offers what it calls "Vary Region and Remix Prompting," which allow users to select and regenerate regions of an image with a modified prompt. In the "Getting Started" section of its website, Midjourney provides the following images to demonstrate how these tools work.

Arranging1

The image was further modified by repeating the editing process:

Arranging2

Other generative AI systems also offer tools that similarly allow users to exert control over the selection, arrangement, and content of the final output.

Unlike prompts alone, these tools can enable the user to control the selection and placement of individual creative elements. Whether such modifications rise to the minimum standard of originality required under Feist will depend on a case-by-case determination. In those cases where they do, the output should be copyrightable.

Similarly, the inclusion of elements of AI-generated content in a larger human-authored work does not affect the copyrightability of the larger human-authored work as a whole. For example, a film that includes AI-generated special effects or background artwork is copyrightable, even if the AI effects and artwork separately are not.

...

C.AIシステムの補助的利用

(略)

アメリカ著作権局は、作品の創作における補助ツールとしてAIを使う事と人間の創作性に代わるものとしてAIを使う事の間には重大な区別があるという事に同意する。人間の表現を強化する補助的利用は著作権保護を制限するものとならない一方で、AIシステムが表現の選択をする利用はさらなる分析を要する。この区別はどの様にシステムが使われているかにより、その本質的な性質によるものではない。

(略)

D.プロンプト

(略)

2.分析

アメリカ著作権局は、一般的に入手可能な現在の技術を前提として、プロンプトのみではAIシステムの利用者をその出力の著作者とするに十分な人間によるコントロールを提供しないと結論づける。プロンプトは保護され得ないアイデアを伝える指示として本質的に機能するものである。高度に詳細化されたプロンプトは利用者の望む表現的要素を含み得るであろうが、今の所それはAIシステムが出力を生成する際にそれを処理するやり方をコントロールしていない。

共同著作に関するケースもこの結論を指示する。これらのケースは著作者である事を主張するのに必要となるコントロールの多寡を論点とするものである。詳細な指示の提供は、どの様にこれらの指示が実行されたかに影響されず、不十分である。第3巡回控訴裁が説明している様に、ある者がその表現を実行するために誰かを雇う時、委託者が最終的な作品における著作権を有する著作者である事を主張するためには「このプロセスが知的な修正又は高度な技術的強化を必要としない機械的な転写」でなければならない。(訳注:原注によると、この記載は1991年5月1日のアンドリアン対南オーシャンカントリー商工会議所事件第3巡回控訴裁判決によるもの。)

生成AIにプロンプトを入力する事は作品を創作する事を委任された芸術家に指示を与える事に似ていると見られ得るが、そこには大きな違いがある。人間対人間の協力作業においては、依頼者側も委託された芸術家の寄与を監督し、指示し、理解する事ができる。それぞれの者の寄与の性質に依存し、芸術家が唯一の著作者である事もあれば、得られた物は共同著作物か依頼のために作られた著作物かである事もある。理論的には、AIシステムもいつの日か、そのシステムの寄与が機械的なものとなるほど、どの様にその表現が出力に反映されるのかに関するコントロールが大いに人にできる様になるかも知れない。今日のAIシステムの動作に関する証拠は今の所そうではな事を示している。プロンプトは作り出される表現的要素を適切に決定しているか、どの様にシステムがそれを出力に翻訳しているかをコントロールしている様には見えない。

プロンプトと結果となる出力の間のギャップは、利用者がそのアイデアの固定された表現への変換におけるコントロールを有しておず、システムが出力における表現的要素を決める上で大きな寄与をしている事を示している。言い換えると、プロンプトは利用者の精神的認識又はアイデアを反映し得るが、利用者はそのアイデアが表現されるやり方をコントロールできない。この事は生成AIシステムが内部的にプロンプトを修正又は書き換えている場合より明確なものとなる。

アメリカ著作権局がプロンプトを商業的に入手可能なAIシステムに入力する事により生成した画像がこの事を示す良い例となるだろう:(訳注:原注によると、使われたのはグーグルのGEMINI。)

Prompt

このプロンプトは望む出力の内容、場面の設定、画像のスタイル、主題の場所を記載している。出力画像はこれらの指示の幾つかを反映している(例えば、パイプを吹かし、眼鏡を掛けた猫)しかし、他はそうではない(例えば、高度に詳細化された木材の環境)。指示が提供されていない所ではAIシステムがそのギャップを埋めている。

例えば、プロンプトは、猫の種別や色合い、サイズ、ポーズ、その特徴や表情のあらゆる要素、又は、それが、もしあれば、ローブの下にどの様な服を着るべきかについて特定していない。不釣り合いな人間の手によって新聞が持たれるべき事をプロンプトは何ら指示していない。

同じプロンプトが違う出力を生成し得る事も人間のコントロールの欠如をさらに示すものである。ある良く使われているシステムがそのウェブサイトで「どの様に記載されるかにかかわらず・・・同じテキストの記載によって無数のあり得る」出力がある事を説明している通り、AIシステムのブラックボックスが利用者の指示の様々に変わる解釈を提供している。(訳注:原注によると、これはAIサービスの1つUdioの説明。)

プロンプトを繰り返し再修正する事もこの分析を変える事はなく、出力について著作権を主張する十分な基礎を提供しない。第一に、著作権が独創的な著作者性を保護するものであって、重労働又は「額の汗」を保護するものではない以上、プロンプトの再修正により作品を製作する際に要した時間、費用又は努力は無関係である。(訳注:原注によると、「額の汗」の記載は1991年5月27日のファイスト出版対ルーラル電話サービス最高裁判決によるもの。)第二に、再修正したプロンプトを入力する事は単一のプロンプトを入力する事と動作において具体的に異なるものとは見えない。何度もプロンプトを再修正して入力する事で、利用者はサイコロを「繰り返し振り」、そこから選択するさらなる出力をシステムに生成させているが、そのプロセスに対するコントロールの程度を変えているという事はない。何回プロンプトが再修正され再入力されるかにかかわらず、最終的な出力は、それが含む表現の著作物性というより、AIシステムの解釈の利用者の受容を反映している。

(略)

E.表現的入力

例として、アメリカ著作権局に登録のために提出された次の作品で、著作者は人手のイラストを作り、それを下に示されるプロンプトとともに入力として使った:(訳注:原注によると、これはクリス・カシュタノヴァ氏の「薔薇の謎」という作品、同氏のHPの説明も参照。)

AIシステムは次の出力を作り出した:

Expressive

イラスト自体が著作権の保護を受けられる作品であり、仮面の輪郭、仮面の形状における鼻、口及び頬骨の位置、茎と薔薇の花弁の配置並びに4つの葉の形状及び場所などのその表現的要素は明らかに出力においても認識できる。

出願人は、鼻、唇及び薔薇の花弁のリアルな3次元の描写並びに背景のライティング及び影などの昨秋的な作品にあらわれた「あらゆる人間でないものによる表現」を除いた。出願において提供された情報を審査し、アメリカ著作権局は次の注記とともに作品を登録した:「登録は、出願において明らかに認識可能であり、クレームから除外されている人間によらないものによる表現から分離可能な、変化を受けていない人間による絵の著作者性に限られる。」(訳注:なお、原注にある通り、別事件として、人間による写真を入力としているが、出力に対する十分な創作的コントロールをしておらず、元の作品の著作物保護適格性も不明として、著作権登録が拒絶されたものもある。2023年12月11日のアメリカ著作権局再審査部決定(pdf)参照。)

この例でより良く示されている通り、人間が著作権保護適格性を有する自らの作品を入力し、その作品が出力において認識できる場合、その者は少なくともその出力のその部分に対する著作者となるであろう。その自身の創作的表現は、派生著作物におけるものと同様の範囲において、著作権による保護を受けられるであろう。派生著作物の保護が後の著作者によって追加されたマテリアルに限られる様に、この種のAI生成出力における著作権は認識可能な人間の表現をカバーするものであろう。それは単独のAI生成要素に及ぶ事はないであろうが、人間の著作になる及びAIにより生成されたマテリアルの選択、調整及び配置もカバーし得るであろう。

F.AI生成物の修正やアレンジ

(略)

多くの良く使われているAIプラットフォームは、利用者にAI生成コンテンツを反復可能な形で選択、編集及び脚色する事を促すツールを提供している。例えば、Midjourneyは、画像の区域を選択し、修正プロンプトとともに再生成する事を利用者に可能とする「区域の変更及びリミックスプロンプト入力」と呼ばれるものを提供している。そのウェブサイトの「始めてみよう」のセクションにおいて、Midjourneyは、このツールがどの様に動くのかを示す次の画像を提供している。(訳注:原注によると、画像下の記載はアメリカ著作権局により追加されたものであり、(1)プロンプトによる候補画像の生成、(2)選択と画像の拡大、(3)フリーハンド編集ツールを用いた区域の選択、(4)プロンプトによる候補画像の生成、(5)選択と画像の拡大という手順を説明している。MidjourneyのHPの説明も参照。)

Arranging1

この画像は編集プロセスを繰り返す事によりさらに次の様に修正される:

Arranging2

他の生成AIシステムも、利用者に選択、アレンジ及び最終的な出力内容に関するコントっロールの実行を同様に可能とするツールを提供している。

プロンプトだけとは異なり、これらのツールは個々の創作的要素の選択及び場所についてのコントロールを利用者に可能としている。この様な修正がファイスト最高裁判例の下で必要とされる独創性の最低限の水準を満たすかどうかはケースバイケースの判断によるであろう。そうと判断されるケースにおいて、出力は著作権の保護を受ける。

同様に、より大きな人間の著作になる作品にAI生成コンテンツの要素が含まれる事はそのより大きな人間の著作になる作品が全体として著作権の保護を受けられる事に影響しない。例えば、AIのエフェクト及び作品が個別には著作権の保護を受けられないとしても、AIにより生成された特別なエフェクトや背景作品を含む映画は著作権の保護を受けられる。

(略)

 アメリカ著作権局は、上の部分で、著作権法の原則に従い、作品の創作にAIを利用した場合において、人間によって創作された表現と言える部分のみが著作権による保護を受けられる事をはっきりと示している。

 この事は2023年3月に公表されたAI生成物の著作権登録を不可とする方針ペーパー(第476回参照)でも書かれていた事であるが、補助的利用、プロンプト入力、表現的入力、出力のアレンジとパターン分けした上で、それぞれに具体的な例もあげられ、非常に分かりやすくまとまっていると言える。

 この様な報告書を読むにあたっては訴訟に必要なために著作権登録が良く利用されているというアメリカ特有の事情を踏まえておく必要があるが、ここで著作権保護適格性について書かれている事は人間の創作を保護するという著作権法の最も根本的な原則からのみ導かれるものであって、どの国でも通用するものである。その意味において、日本の文化庁による生成AIと著作権の関係に関する報告書ともその考え方は共通すると言って良い。(文化庁の報告書については第492回、他の知財権との関係も含む知財本部の報告書については第495回参照。)

 第Ⅲ章の「国際動向」も、上記の日本の報告書に対する言及なども含め世界各国の状況について簡潔明瞭にかかれていて興味深いが、最後の第Ⅳ章「法改正に関する意見」で、以下の通り、AI生成物に対する追加的な保護を目的とした法改正について否定的な意見を述べているのは特に注目に値する。

A. Providing Incentives

...

In the Office's view, the case has not been made for additional protection for AI-generated material beyond that provided by existing law. As an initial matter, because copyright requires human authorship, copyright law cannot be the basis of protection for works that do not satisfy that requirement. As most commenters recognized, the incentives authorized by the Copyright Clause are to be provided to human authors as the means to promote progress. While Congress could instead consider establishing sui generis rights, we do not find the policy arguments for additional protection to be persuasive.

To begin with, it is not clear that new incentives are needed. The developers of AI models and systems already enjoy meaningful incentives under existing law (as indicated by the rapid development and adoption of those models and systems). These incentives include patent, copyright, and trade-secret protection for the machinery and software, as well as potential funding and first-mover advantages. Moreover, we are not convinced that providing further incentives would promote progress.

We share the concerns expressed about the impact of AI-generated material on human authors and the value that their creative expression provides to society. If a flood of easily and rapidly AI-generated content drowns out human-authored works in the marketplace, additional legal protection would undermine rather than advance the goals of the copyright system. The availability of vastly more works to choose from could actually make it harder to find inspiring or enlightening content. Indeed, AI training itself is reportedly reliant on human-generated content, with synthetic data leading to lower-quality results.

...

D. Providing Greater Clarity

...

The Office understands the desire for clarity around the copyrightability of AI-generated material. We do not believe, however, that legislation is necessary at this point. Much of the concern expressed focused on the assistive use of AI tools, and this Report seeks to provide assurances that such uses do not undermine protection. As to determining the copyrightability of AI outputs, the courts will provide further guidance on the human authorship requirement as it applies to specific uses of AI (including in reviewing the Office's registration decisions).

Meanwhile, the analysis in this Part of the Report can help to shed light on how existing principles and policies apply.

Even if Congress were to consider addressing this issue through legislation, greater clarity would be difficult to achieve. Because the copyrightability inquiry requires analysis of each work and the context of its creation, statutory language would be limited in its ability to provide brighter lines. Unless and until future developments raise new problems, the Office does not recommend a change in the law.

...

A.インセンティブの提供

(略)

アメリカ著作権局の見解としては、既存の法によって提供されるものを超えてAI生成マテリアルに対する追加の保護をするべき事実はない。最初の事として、著作権は人間の著作者を必要とするのであって、著作権法はこの要件を満たさない作品を保護する基礎となり得ない。ほとんどの意見が認める通り、アメリカ憲法の著作権条項により正当化されるインセンティブは、発展を促進する手段として人間の著作者に提供されるものである。議会はその代わりに独自の権利を創設する事を検討できるであろうが、追加の保護に関する政策的主張が説得力を有するとは私たちには思えない。

まず、新たなインセンティブが必要とされる事が明確でない。AIモデル及びシステムの開発者は(これらのモデル及びシステムの急速な発展及び採用によって示される通り)既に既存の法の下で有意なインセンティブを享受している。これらのインセンティブは、機械及びソフトウェアに関する特許、著作権及び営業秘密保護並びに可能な資金調達及び先行者の利益を含む。また、さらなるインセンティブが発展を促進するであろうとは私たちには思えない。

人間の著作者及びその創作的表現が社会に与える価値に対するAI生成マテリアルの与える影響に関して表明された懸念を私たちは共有する。もし容易に素早くAIにより生成されたコンテンツが市場に溢れ、人間により著作された作品を押し流す事があったとして、追加の保護は著作権制度の目的に適うものというよりそれに反するものとなるであろう。そこから選択可能なより莫大な作品の入手可能性は見事な又は有意義なコンテンツを見つける事を実際困難にし得るであろう。実際、AI学習それ自体は人間により作り出されたコンテンツに依拠している事が報告されており、合成データでは低品質の結果がもたらされる。

(略)

D.明確化の提供

(略)

アメリカ著作権局は、AI生成マテリアルの著作権による保護の可能性を巡り明確性を求める要望を理解する。しかしながら、私たちはこの点において立法が必要とは思わない。表明された懸念の多くはAIツールの補助的利用に着目したものであり、この報告書はその様な利用は保護を弱めるものではないという安心を提供しようとするものである。AI出力の著作権による保護の可能性の判断に関して、(著作権局の登録に関する決定の再検討を含め)特定のAI利用に適用される場合の人間による著作者性の要件についてさらなる指針は裁判所により提供されるであろう。

それと同時に、この報告書のこの部分の分析は既存の原則及び方針がどの様に適用されるかについて光を投げかける助けとなり得るであろう。

たとえアメリカ議会がこの問題について立法による対処を検討したとしても、より明確化を図る事を達成するのは難しいであろう。著作権による保護に関する検討はそれぞれの作品とその創作の文脈の分析を必要とするのであって、法律の文言はより明確な線を引くその能力において限定的であろうからである。今後の開発により新たな問題が生じない限り、また、それまで、アメリカ著作権局は法律の改正を推奨しない。

(略)

 日本でもAI生成物に知的財産の側面から追加の保護を与える事を考えても良いとする意見がある様に思うが、私自身は全く否定的であり、その理由はここで書かれている事にほぼ尽きている。

 純粋なAI生成物に追加の保護を与える事は、文化の発展を目的として人の創作を保護して人にそのインセンティブを与えるという著作権法の根本原則に真っ向から反する事に他ならず、機械であるAIに対して新たなインセンティブを導入する理由は全くない。また、その様な事を検討したとしても、立法技術的に困難というのもこの報告書で書かれている通りだろう。

 さらに書いておくと、この事は特許法による保護についても該当する。この報告書でも書かれている通り、人が考え出したAIを利用した発明が特許を受けられ、その点で既にAI技術の開発のためのインセンティブは与えられている。繰り返しになるが、著作権法からであれ、特許法からであれ、人の創作を保護する創作保護法としての知的財産法の観点から、何かしらの立法により機械であるAIに対して新たなインセンティブを導入する理由は何一つないのである。

 一番最初に書いた通り、このアメリカ著作権局の報告書は特に新しい事が書かれているというものではないが、著作権法の根本的な原則に基づき、AIを利用した場合の著作権保護適格性の考え方を、また、AI生成物の追加的な保護の立法に対する否定的な見解を非常に明確に示したものとして一読の価値があるものであって、AIと著作権の問題に関する国際動向の1つとして日本の議論においても当然参照されるべきものである。

 上でリンクを張ったリリースにも書かれている通り、このアメリカ著作権局の報告書は3部構成で、最後の第3部はAI学習に関する予定との事であり、それも公表され次第また紹介したいと思っている。

(2025年5月18日の追記:タイトルにどこの報告書かと第何部かをつけた方が良いと思ったので修正し(「第508回:・・・報告書」→「第508回:・・・アメリカ著作権局の報告書第2部」)、合わせて第1段落も直した(「報告書」→「報告書の第2部」)。)

| | コメント (0)

2024年9月 1日 (日)

第501回:人口知能(AI)によるデジタルレプリカ(ディープフェイク)に関するアメリカ著作権局の報告書

 先月末、7月31日にアメリカ著作権局が人口知能(AI)関係報告書の第1部として、デジタルレプリカに関する報告書を出しているので、今回はこの報告書について取り上げる。

 この報告書第1部(pdf)は、米著作権局のリリースに、幾つかのパートに分かれて出されるものの第1部であって、AIと著作権法の関係に関する主たる問題である、生成AIによって生成されたマテリアルの著作物性、著作物に基づいて行われるAI学習の法的評価、ライセンスの考慮、あり得る責任の所在といった事項は全て次のパート以降に持ち越しとなっている。

 日本ではそれほど使われていない様に思うが、この報告書で言う所のデジタルレプリカとは、「Ⅰ イントロダクション」の第2ページで、

This Report uses the term "digital replica" to refer to a video, image, or audio recording that has been digitally created or manipulated to realistically but falsely depict an individual. A "digital replica" may be authorized or unauthorized and can be produced by any type of digital technology, not just AI. The terms "digital replicas" and "deepfakes" are used here interchangeably.

本報告書では、「デジタルレプリカ」という用語は、リアルだが偽って個人を描き出す様にデジタル的に作られたか操作されたものである動画、画像又は音声レコーディングの事を指す。「デジタルレプリカ」は許諾を受けている事もあれば許諾を受けていない事もあり、AIだけでなくあらゆるタイプのデジタル技術によって作り出され得る。デジタル用語の「デジタルレプリカ」と「ディープフェイク」はここで交換可能なものとして使われる。

と書かれている通り、実質的に個人を対象としてリアルに作られた偽情報を意味しており、日本でもそのまま使われている語としてはディープフェイクとほぼ同じ意味と考えて良いものである。

 話としては少し前後するが、この様にディープフェイク、偽情報対策を主眼に置いている事から、この報告書と著作権法との関係は必然的に薄くなっている。この点については、各法との関係を記載している部分の第17ページで、

Copyright protects original works of authorship, including the material - photographs or audio or video recordings - from which a digital replica might be constructed. The Copyright Act provides copyright owners with a bundle of exclusive rights, including the rights to reproduce a work and to prepare derivative works.

Digital replicas that are produced by ingesting copies of preexisting copyrighted works, or by altering them - such as superimposing someone's face onto an audiovisual work or simulating their voice singing the lyrics of a musical work - may implicate those exclusive rights. If the depicted individual is an owner of the copyrighted work, he or she could have a copyright claim for infringement of the work as a whole. Copyright does not, however, protect an individual's identity in itself, even when incorporated into a work of authorship. A replica of their image or voice alone would not constitute copyright infringement.

著作権は、デジタルレプリカががそこから作られ得るであろうマテリアル-写真、音声又は動画のレコーディングなど-を含む、著作者による独創的な著作物を保護するものである。著作権法は、著作権者が著作物を複製し、派生著作物を用意する権利を含む権利の束を持つ事を規定している。

既存の著作物のコピーを取り込むか、-ある者の顔を映像作品に入れ込む、その声に音楽作品の歌詞を歌わせる様にシミュレートするなど-それを作り変えるかする事によって事によって作り出されるデジタルレプリカは、この排他的権利と関係し得るものである。もし描き出された個人が著作物の権利者であれば、彼又は彼女はその著作物全体の侵害に対して著作権に基づく主張をし得るであろう。しかしながら、著作権は、それが著作者による著作物に化体している場合でも、個人のアイデンティティそれ自体を守るものではない。その画像又は声のレプリカだけでは著作権侵害を構成しないであろう。

と書かれている通りであって、著作権はあくまで創作的表現を守るものであって個人のアイデンティティそれ自体を守るものではないのである。

 さらに、著作権との関係については、第53ページからの「Ⅲ 芸術的スタイルの保護」で書かれている事もあるが、ここでも、その後半の第54~55ページで、

The Office acknowledges the seriousness of these concerns and believes that appropriate remedies should be available for this type of harm.

Copyright law's application in this area is limited, as it does not protect artistic style as a separate element of a work. As noted by several commenters, copyright protection for style would be inconsistent with section 102(b)'s idea/expression dichotomy. Moreover, in most cases the elements of an artist's style cannot easily be delineated and defined separately from a particular underlying work. Google and EFF both stressed that, as a policy matter, stylistic aspects of expressive content should remain freely available for later creators to develop and build on.

The Copyright Act may, however, provide a remedy where the output of an "in the style of" request ends up replicating not just the artist's style but protectible elements of a particular work. Additionally, as future Parts of this Report will discuss, there may be situations where the use of an artist's own works to train AI systems to produce material imitating their style can support an infringement claim.

Numerous commenters pointed out that meaningful protections against imitations of style may be found in other legal frameworks, including the Lanham Act's prohibitions on passing off and unfair competition. In its comments, the FTC stated:

[M]imicking the creator's writing style ... may also constitute an unfair method of competition or an unfair or deceptive practice, especially when the copyright violation deceives consumers, exploits a creator's reputation or diminishes the value of her existing or future works, reveals private information, or otherwise causes substantial injury to consumers.

著作権局は、これらの懸念の深刻さを認め、その適切な救済措置がこの種の害に対して入手可能とされるべきであると考える。

それが芸術的なスタイルを著作物から分離される要素として保護をしていない以上、この分野における著作権法の適用は限られている。何人かの意見で言われている通り、スタイルに対する著作権の保護は、アメリカ著作権法第102条(b)のアイデア/表現二分論と合致しないであろう。さらに、ほとんどの場合において、芸術家のスタイルの要素は、元の特定の著作物から分けて簡単に線引きし、分離する事はできない。グーグルとEFFはともに、政策的事項として、表現物のスタイル的な面は後の創作者が発展させ、積み上げていけるよう自由に利用可能な儘とされるべきと強調している。

しかしながら、著作権法はスタイルにおける要求の結果が芸術家のスタイルだけではなく特定の著作物の保護され得る要素も複製するに至った場合の救済措置を提供している。また、この報告書の将来の部分で検討する予定であるが、そのスタイルを真似るマテリアルを作り出すためにAIシステムを訓練するための芸術家の自身の著作物の利用が、侵害主張の支えとなる状況はあり得るであろう。

多くの意見で、スタイルの模倣に対する有意な保護は、パッシングオフ(詐称通用)及び不正競争に対するランハム法の禁止を含む他の法的枠組みに見つかると指摘されている。その意見において、連邦取引委員会は以下の様に述べている:

創作者の文体の真似も…特に、著作権侵害が消費者を騙すか、創作者の評判を利用するか、その既存の又は将来の著作物の価値を損なうか、個人情報を晒すか、その他消費者に対して実質的な害をなすかする場合に、競争における不公正な方法、不公正な又は欺瞞的な行為を構成し得る。

と、アイデア表現二分論から、表現とは分離された要素としてのスタイル(画風や文体など)に著作権法の保護は原則及ばないという極当たり前の事が書かれ、表現として保護され得る様な状況については今後の報告書で書く予定としているのである。

(ここで詳細に論じる事はしないが、アイデア表現二分論とは、著作権法で守られるのは著作物における表現であって、そこに含まれるアイデアには著作権法の保護は及ばないとする考え方の事である。アメリカ著作権法第102条(b)の様にわざわざそのための明文の規定があるといった事はないが、これは著作権法の世界における基本的な考え方の1つであって、日本の著作権法でも同様に通用する。なお、日本におけるAIと著作権法の関係整理については第492回参照。)

 それでは何が書かれているかというと、上で翻訳した部分でも他の法律について少し書かれているが、要するに、第8ページからの「Ⅱ 許諾を得ていないデジタルレプリカに対する保護」の「A.既存の法的枠組み」で、以下の様に、アメリカのコモンロー、州法、連邦法でデジタルレプリカ又はディープフェイクに適用可能なものをあげ、全て一長一短ある事から、統一的な連邦法の制定による対応が望ましいという事が言われているのである。

  1. 各州のコモンロー及び州法(第8ページ~)
    a)プライバシーの権利(第8ページ~、日本では通常個人がその私生活に干渉されない事を言うと思うが、アメリカでは偽情報の流布による個人の評判に対する重大な危害に対する保護も含まれる)
    b)パブリシティの権利(第10ページ~、日本と同様、アメリカでもパブリシティ権は著名な者に対してその名称等の商業的な利用を保護する)
    c)デジタルレプリカに対する州による新規制(第15ページ~、テネシー州、ルイジアナ州、ニューヨーク州の新法を紹介)
  2. 連邦法(第16ページ~)
    a)著作権法(第17ページ~、ただし、上で訳出した通り、デジタルレプリカと著作権法の関係は薄い)
    b)連邦取引委員会法(第17ページ~、当然の事だが、AI技術の反競争的な利用に対しては競争法の適用が考えられる)
    c)ランハム法(第19ページ~、ランハム法とは商標法の事だが、これは一部日本で言う所の不正競争防止法的側面を含み、登録商標に基づく商標権侵害だけでなく一般的に偽の表示による出所の混同なども規制している)
    d)通信法(第20ページ~、消費者を対象とする電話を規制している電話消費者保護法の適用が考えられる)
  3. 私的取り決め(第21ページ~、私人間のその名前等の利用に関する契約もあり得る)

 そして、その後の「B.連邦立法の必要性」で、各州のコモンローや州法の規定や解釈にはばらつきがある事、アメリカの国会でも幾つかの法案が提出される等しているが、その内容にかなりの差異があり、議論はまだ煮詰まっていない事、新しい法律における権利の考慮要素としてa)規制対象、b)保護を受ける者、c)保護期間、d)侵害となる行為((i)商業的利用に限られるか、(ⅱ)知っている事を必要とするか、(ⅲ)2次的責任(間接侵害)についてどの様に整理するか)、e)ライセンスや譲渡についてどうするか((ⅰ)期間について、(ⅱ)インフォームドコンセントについて(ⅲ)未成年との契約について)、f)アメリカ憲法修正第1条(表現の自由)に基づく懸念についてどの様に整理するか、g)救済措置、f)連邦法の優先(連邦法と矛盾する州法を無効とする法理)がある事をあげ、最後に第57ページの結論で、以下の通り締め括っている。

The Copyright Office agrees with the numerous commenters that have asserted an urgent need for new protection at the federal level. The widespread availability of generative AI tools that make it easy to create digital replicas of individuals' images and voices has highlighted gaps in existing laws and raised concerns about the harms that can be inflicted by unauthorized uses.

We recommend that Congress establish a federal right that protects all individuals during their lifetimes from the knowing distribution of unauthorized digital replicas. The right should be licensable, subject to guardrails, but not assignable, with effective remedies including monetary damages and injunctive relief. Traditional rules of secondary liability should apply, but with an appropriately conditioned safe harbor for OSPs. The law should contain explicit First Amendment accommodations. Finally, in recognition of well-developed state rights of publicity, we recommend against full preemption of state laws.

The Office remains available as a resource to Congress, the courts, and the executive branch in considering the recommendations in this Report and future developments.

著作権局は、連邦レベルでの新たな保護が喫緊で必要である事を主張する多くの意見提出者に同意する。広く利用可能となっている、個人の画像と動画のデジタルレプリカを容易に作る事ができる生成AIツールは、既存の法律と許諾を得ていない利用によって加えられ得る害について持ち上がっている懸念の間のギャップを浮き彫りにしている。

私たちは、許諾を得ていないデジタルレプリカのそうと知っての頒布から全ての個人をその生涯を通じて守る権利を議会が確立する事を推奨する。この権利はガードレール内でライセンス可能であるが、譲渡不可能であり、金銭的損害賠償と差し止めを含む有効な救済措置を伴うべきでる。2次的責任に関する伝統的な規則が適用されるべきであるが、オンラインサービスプロバイダーに対する適切に条件が設けられたセーフハーバーを伴うべきである。法は憲法修正第1条との明文の調整規定を含むべきである。最後に、十分発展している各州のパブリシティの権利を認め、私たちは州法に対する連邦法の完全な優先に反対の立場を取る事を勧める。

著作権局は、本報告書における推奨及び将来的な展開を考慮するに際し、議会、裁判所、行政機関に対してリソースとしてさらに利用可能である。

 上でも書いた通り、この米著作権局の報告書は、著作権法との関係は薄く、各州におけるコモンローや州法といったものがなくパブリシティ権等も最高裁の判例によって認めらている日本との関係で直ちに参考になるといったものではないが、デジタルレプリカ(ディープフェイク)に対する個人の保護に関するアメリカにおける現在の議論を簡潔に分かり易くにまとめたものとして非常に興味深いものである。この様な著作権法以外の法律を中心として立法に関してもかなり踏み込んだ見解が報告書が先にまとめられた背景には、米著作権局のパブリックコメントで著作権法以外の法律との関係が主として問題となるデジタルレプリカに対する懸念が多く提起されていた事、著作権局が議会の付属機関である事もあるだろう。

 また、最近のアメリカにおけるAI規制に関する動きとしては、日本でも報道されている様に(時事通信の記事や、朝日新聞の記事参照)、8月末にカリフォルニア州の議会を通過し、今現在州知事の署名を待つ状態になっている、1億ドル以上の開発費のAIモデルに対して安全対策を課すものであるカリフォルニア州法の先端人口知能モデルのための安全安心なイノベーション法がある。

 この儘知事の署名を得て成立し、IT企業が集中するカリフォルニア州の州法として一定の影響を与えて行く事になるかも知れないが、この州法のレベルでも賛否両論が吹き出している事には大いに留意すべきだろう。

 どこの国においてもAIに関する問題として本当に中心として議論すべきは、なぜか真っ先に取沙汰されがちな知的財産法、著作権法との関係ではなく、この様な偽情報対策だろうと私は常に思っている。ただ、上の米著作権局の報告書からも連邦レベルでの議論はアメリカでもまだ十分に煮詰まってない事が見て取れ、今年はアメリカの選挙年でもあり、本当の意味で国としてのアメリカのAI規制論の方向性がはっきりと出て来るのはもう少し先の事になるのではないかと私は見ている。

(2025年2月9日の追記:1箇所誤記を直した(「による報告書」→「関係報告書」)。)

| | コメント (0)

2023年9月 3日 (日)

第483回:人工知能(AI)は著作者たり得ないとするアメリカ地裁の判決とアメリカ著作権局のAIに関する意見募集

 今回は、人工知能(AI)に関し、参考としてアメリカの動きを2つ取り上げる。

(1)AIは著作者たり得ないとするアメリカ地裁の判決
 アメリカでもAIと著作権の問題に関するものとしてはほぼ最初のものだろうと思うが、この8月18日にアメリカのコロンビア地裁が、AIは著作者たり得ないとする判決(pdf)を出した。

 このターラー対パールムッター事件は、第476回で取り上げたアメリカ著作権局の方針ペーパーでも言及されていたものであり、判決のポイントとなる部分を以下に訳出するが、コロンビア地裁はAI生成物は著作権登録ができないとするアメリカ著作権局の判断を全面的に支持した。(以下、いつも通り翻訳は全て拙訳である。)

... Human involvement in, and ultimate creative control over, the work at issue was key to the conclusion that the new type of work fell within the bounds of copyright.

Copyright has never stretched so far, however, as to protect works generated by new forms of technology operating absent any guiding human hand, as plaintiff urges here. Human authorship is a bedrock requirement of copyright.

...

The 1976 Act's "authorship" requirement as presumptively being human rests on centuries of settled understanding. The Constitution enables the enactment of copyright and patent law by granting Congress the authority to "promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries." ...

The human authorship requirement has also been consistently recognized by the Supreme Court when called upon to interpret the copyright law. As already noted, in Sarony, the Court's recognition of the copyrightability of a photograph rested on the fact that the human creator, not the camera, conceived of and designed the image and then used the camera to capture the image. ... In all these cases, authorship centers on acts of human creativity.

Accordingly, courts have uniformly declined to recognize copyright in works created absent any human involvement, even when, for example, the claimed author was divine. The Ninth Circuit, when confronted with a book "claimed to embody the words of celestial beings rather than human beings," concluded that "some element of human creativity must have occurred in order for the Book to be copyrightable," for "it is not creations of divine beings that the copyright laws were intended to protect." Urantia Found. v. Kristen Maaherra, 114 F.3d 955, 958-59 (9th Cir. 1997) ... Similarly, in Kelley v. Chicago Park District, the Seventh Circuit refused to "recognize[] copyright" in a cultivated garden, as doing so would "press[] too hard on the[] basic principle[]" that "[a]uthors of copyrightable works must be human." 635 F.3d 290, 304-06 (7th Cir. 2011). The garden "ow[ed] [its] form to the forces of nature," even if a human had originated the plan for the "initial arrangement of the plants," and as such lay outside the bounds of copyright. Id. at 304. Finally, in Naruto v. Slater, the Ninth Circuit held that a crested macaque could not sue under the Copyright Act for the alleged infringement of photographs this monkey had taken of himself, for "all animals, since they are not human" lacked statutory standing under the Act. 888 F.3d 418, 420 (9th Cir. 2018). ... Plaintiff can point to no case in which a court has recognized copyright in a work originating with a non-human.

Undoubtedly, we are approaching new frontiers in copyright as artists put AI in their toolbox to be used in the generation of new visual and other artistic works. The increased attenuation of human creativity from the actual generation of the final work will prompt challenging questions regarding how much human input is necessary to qualify the user of an AI system as an "author" of a generated work, the scope of the protection obtained over the resultant image, how to assess the originality of AI-generated works where the systems may have been trained on unknown pre-existing works, how copyright might best be used to incentivize creative works involving AI, and more. See, e.g., Letter from Senators Thom Tillis and Chris Coons to Kathi Vidal, Under Secretary of Commerce for Intellectual Property and Director of the U.S. Patent and Trademark Office, and Shira Perlmutter, Register of Copyrights and Director of the U.S. Copyright Office (Oct. 27, 2022), https://www.copyright.gov/laws/hearings/Letter-to-USPTO-USCO-on-National-Commission-on-AI-1.pdf (requesting that the United States Patent and Trademark Office and the United States Copyright Office "jointly establish a national commission on AI" to assess, among other topics, how intellectual property law may best "incentivize future AI related innovations and creations").

This case, however, is not nearly so complex. While plaintiff attempts to transform the issue presented here, by asserting new facts that he "provided instructions and directed his AI to create the Work," that "the AI is entirely controlled by [him]," and that "the AI only operates at [his] direction," Pl.'s Mem. at 36-37 - implying that he played a controlling role in generating the work - these statements directly contradict the administrative record. Judicial review of a final agency action under the APA is limited to the administrative record ... Here, plaintiff informed the Register that the work was "[c]reated autonomously by machine," and that his claim to the copyright was only based on the fact of his "[o]wnership of the machine." Application at 2. The Register therefore made her decision based on the fact the application presented that plaintiff played no role in using the AI to generate the work, which plaintiff never attempted to correct. ... Plaintiff's effort to update and modify the facts for judicial review on an APA claim is too late. On the record designed by plaintiff from the outset of his application for copyright registration, this case presents only the question of whether a work generated autonomously by a computer system is eligible for copyright. In the absence of any human involvement in the creation of the work, the clear and straightforward answer is the one given by the Register: No.

...

(略)問題となる作品における人間の関与、そして、それに対する究極的な創作的コントロール、が著作物の新しいタイプが著作権の領域内に入るという結論を導く鍵だった。

しかしながら、ここで原告が主張しているように、著作権は、人間の手による導き抜きの技術的な実施の新しい形式によって生成された作品の保護まで拡張される事は決してなかった。人間の著作者が著作権の根本要件である。

(略)

人間である事と考えられる1976年著作権法の「著作者」要件は何世紀にも渡り安定した理解により続いている。憲法は、「著作者及び発明者に限られた期間そのそれぞれの著作と発見に関して排他権を保証する事により、科学及び有益な技芸を促進し、発展させる」権限を議会に与える事により、著作権及び特許法の法制化を可能としている。(略)

人間の著作者要件は著作権法の解釈が求められる場合最高裁によっても一貫して認められて来ている。既に記載した通り、1884年3月17日ののサロニー事件判決において、最高裁の写真の著作物性の認定は、カメラではなく、人間の創作者が、画像を着想し、デザインし、そして、カメラを使って画像を撮影するという事実に基づいていた。(略)これらの全ての事件で、著作者性は人間の創造性による行為を中心としている。

その通り、裁判所は一致して人間が関与する事なく創作された作品において著作権を認める事はなかった、例えば、主張された著作者が神聖なものである場合などにおいてである。第9巡回控訴裁は、「人間ではなく天上の存在の言葉を具現化したものと主張された」本に向かった時、その本が著作権の保護を受けるためには何がしかの人間による創作の要素があり」、「著作権法が保護の目的としているのは神聖な存在の創作ではない」と結論づけた。1997年6月10日のウランティア協会対クリステン・マアヘラ事件第9巡回控訴裁判決参照。(略)同様に、2011年2月15日のケリー対シカゴ公園区事件判決において、第7巡回控訴裁は、そうする事は「著作権の保護を受ける作品の著作者は人間である」という「基本的な原理に余りにも反している」として、整えられた庭園に著作権を認める事を拒絶した。「最初の植物の配置」について人間が計画を作り、それを始めたとしても、庭園は「その形式を自然の力に負っている」のである。最後に、2018年4月23日のナルト対スレーター事件判決において、第9巡回控訴裁は、「あらゆる動物は、人間ではないのであるから」著作権法における原告適格性を欠き、クロザルはこの猿が自身を撮った写真の被疑侵害に対して著作権法の下で訴えを起こす事はできないと判断した。(略)原告は裁判所が人間ではないものによって作られた作品に著作権を認めたケースを提示できていない。

疑いなく、私たちは、芸術家たちが新しい視覚的及びその他の芸術作品の生成において使われるそのツールボックスの中にAIを置くという著作権における新しいフロンティアに近づいている。最終作品の実際の生成から人間の創造性が減って行く事は、どれほどの人間の入力がAIシステムの利用者が生成された作品の「著作者」と認められるために必要なのか、得られた画像に対して手に入られる保護の範囲、システムが不明の既存作品により学習されたであろう場合におけるAI生成作品の創造性をどのように評価するか、AIが関係する創造的作品に対してインセンティブを与えるために著作権はどの様に使われるのが最適なのか等々について難しい質問を投げ掛けるだろう。例えば、トム・ティリス及びクリス・クーンズ上院議員からカティ・ヴィダルアメリカ特許庁長官及びシーラ・パールムッターアメリカ著作権局長への手紙、https://www.copyright.gov/laws/hearings/Letter-to-USPTO-USCO-on-National-Commission-on-AI-1.pdf参照(アメリカ特許庁及びアメリカ著作権局に、とりわけ、知的財産法はどの様に最も良く「未来のAI関連発明及び創作にインセンティブを与える」事ができるかを評価するために「ともにAIに関する国レベルの検討委員会を立ち上げる」事を求めている)。

この事件は、しかしながら、それほど複雑ではない。原告は、「指示を与えてそのAIを導き作品を創作した」と、「AIは完全にコントロールされていた」と、そして、「AIは導かれた様に動作しただけである」という新しい事実を主張する事によって、-原告が生成された作品においてコントロールする立場にあった事を示唆して-本件を変容しようと試みているが、この主張は行政記録と直接的に矛盾する。行政手続法の下での省庁の最終的な決定に対する司法審査は行政記録に限られる。(略)ここで、原告は著作権局にその作品は「機械によって自動的に作り出された」ものであり、その著作権に対する請求はその「機械の所有」という事実のみに基づくものであると伝えていた。すなわち、著作権局は、原告が修正しようと試みる事のなかった、その作品を生成するためのAI利用に際し原告は何の役割も果たさなかったという申請において提示されていた事実に基づきその決定をした。行政手続法に基づく司法審査の請求んために事実を更新し、修正しようとする原告の試みは遅きに失している。著作権登録のための申請の当初から原告によって示された記録に基づくと、この事件において問題となるはコンピュータシステムによって自動的に生成された作品は著作権の保護適格性を有するかどうかという質問のみである。作品の創作において人の関与がない場合において、明確かつ単純な答えは著作権局に与えられたものである:否。

(略)

 上で抜き出した部分に書かれている通りだが、AIと著作権の問題に関して、AI生成物に著作権が認められるためにはどれほど人間の関与が必要か、AI学習における既存の著作物の利用をどの様に評価するか、著作権が認められる場合に保護の範囲はどうなるか、AI利用創作との関係で著作権法はどうあるべきかという問題が発生しつつあるというのこの事件を担当したハウエル判事の指摘する通りであって、世界的にもこれらの点が問われているのは間違いない。

 ただし、この事件の判決では、AI生成物に対して原告の指示、管理があったというのは時宜に遅れた主張であって認められず、当初の主張に沿って機械が自動的に生成したものに著作権は認められないとした米著作権局の決定は異論なく正しいとしている事にも注意しておいてよいだろう。ハウエル判事が上で正しく指摘している所のより本質的な質問に対する判断は、控訴審か、恐らく他の事件に持ち越されたのである。(そのため、この判決は先例性のないメモランダムオピニオンとされているのだろう。また、行政訴訟においてどこまで主張や証拠の補足が認められるべきかという事は別の法的論点として非常に興味深いところだが、ここではひとまずおく。)

(2)アメリカ著作権局の著作権とAIに関する意見募集
 もう1つは、8月30日にアメリカ著作権局が、そのリリースの通り、第476回で取り上げた方針ペーパーでも言及されていたものだろう、パブリックコメント・意見募集を開始したという事である。

 その募集要項(pdf)から、意見募集の内容の概要について述べている部分を以下に訳出する。

IV. The Current Inquiry

Drawing on our prior AI Initiative work, including discussions with stakeholders, the Office has identified a wide range of copyright policy issues arising from the development and use of AI. These relate to: (1) the use of copyrighted works to train AI models; (2) the copyrightability of material generated using AI systems; (3) potential liability for infringing works generated using AI systems; and (4) the treatment of generative AI outputs that imitate the identity or style of human artists. The Office seeks public comments on these and related issues.

As to the first issue, the Office is aware that there is disagreement about whether or when the use of copyrighted works to develop datasets for training AI models (in both generative and non-generative systems) is infringing.(Note 34: In some cases, a non-generative AI model may be trained on copyrighted material. In other cases, the same AI model may be capable of being deployed in both a generative AI system and a non-generative one. The Office's consideration of training is framed broadly in order to encompass these and other situations.) This Notice seeks information about the collection and curation of AI datasets, how those datasets are used to train AI models, the sources of materials ingested into training, and whether permission by and/or compensation for copyright owners is or should be required when their works are included. To the extent that commenters believe such permission and/or compensation is necessary, the Office seeks their views on what kind of remuneration system(s) might be feasible and effective. The Office also seeks information regarding the retention of records necessary to identify underlying training materials and the availability of this information to copyright owners and others.

On the second issue, the Office seeks comment on the proper scope of copyright protection for material created using generative AI. Although we believe the law is clear that copyright protection in the United States is limited to works of human authorship,(Note 35: See Mem. Op., Thaler v. Perlmutter, No. 22-cv-1564, ECF No. 24 (D.D.C. Aug. 18, 2023) (affirming the Office's registration denial of AI-generated work).) questions remain about where and how to draw the line between human creation and AI-generated content. For example, are there circumstances where a human's use of a generative AI system could involve sufficient control over the technology, such as through the selection of training materials and multiple iterations of instructions ("prompts"), to result in output that is human-authored? Resolution of this question will affect future registration decisions. While the Office is separately working to update its registration guidance on works that include AI-generated material,(Note 36: For example, the Office has received questions about how to apply its guidance that applicants disclose more than de minimis amounts of AI-generated material in their works. See AI Registration Guidance, 88 FR at 16193 (explaining that "AI-generated content that is more than de minimis should be explicitly excluded from the application").) this Notice explores the broader policy questions related to copyrightability.

On the third question, the Office is interested in how copyright liability principles could apply to material created by generative AI systems.(Note 37: Some of these questions are currently before the courts in lawsuits that have already been filed over generative AI systems. See, e.g., J.L. v. Alphabet Inc., 3:23-cv-03340 (N.D. Cal.); Kadrey v.Meta Platforms, Inc., 3:23-cv-3417 (N.D. Cal.);Silverman v. OpenAI, Inc., 4:23-cv-3416 (N.D. Cal.); Tremblay v. OpenAI, Inc., 3:23-cv-3223 (N.D. Cal.); Getty Images (US), Inc. v. Stability AI, Inc., 1:23-cv-0135 (D. Del.); Andersen v. Stability AI Ltd., 3:23-cv-0201 (N.D. Cal.); Doe v. GitHub, Inc., 4:22-cv-6823 (N.D. Cal.).) For example, if an output is found to be substantially similar to a copyrighted work that was part of the training dataset, and the use does not qualify as fair, how should liability be apportioned between the user whose instructions prompted the output and developers of the system and dataset?

Lastly, in both our listening sessions and other outreach, the Office heard from artists and performers concerned about generative AI systems' ability to mimic their voices, likenesses, or styles. Although these personal attributes are not generally protected by copyright law, their copying may implicate varying state rights of publicity and unfair competition law, as well as have relevance to various international treaty obligations. (Note 38: See U.S. Copyright Office, Authors, Attribution, and Integrity: Examining Moral Rights in the United States 112-116 (Apr. 2019), https://www.copyright.gov/policy/moralrights/full-report.pdf (discussing how such interests are generally protected under state right of publicity laws).)

Ⅳ.本意見募集

利害関係者との議論を含む、私たちの当初のAIイニシアティブの取り組みから、著作権局はAIの発展と利用から生じる広い範囲の著作権政策事項を特定した。これらは次の事に関係する:(1)AIモデルの学習への著作物の利用;(2)AIシステムを利用して生成された物の著作権保護適格性;(3)AIシステムを利用して生成された侵害物についてあり得る責任;(4)人間のアーティストの特性又はスタイルを真似る生成AI出力物の取扱い。著作権局はこれら及び関係する事項についてパブリックコメントを求める。

最初の事項について、著作権局には、AIモデル(生成及び非生成システムの両方で)の学習のためのデータセットの開発への著作物の利用が侵害となるか、どのようなときに侵害となるかについて不一致があることが分かっている。(原注:幾つかの場合において、非生成AIモデルは著作物で学習される事がある。他の場合において、同じAIモデルは生成AIシステム及び非生成AIシステムの両方に配置され得るものである。著作権局の学習に関する考えはこれらの及び他の場合を含むべく広いものである。)本意見募集は、AIモデルの学習のためにこれらのデータセットがどの様に使われるのか、学習に取り込まれる物のソース、そして、その著作物が含まれるとき、著作権者による許可及び/又はのための補償が必要とされるのか又はされるべきかなど、AIデータセットの収集及び整理に関する情報を求めており、著作権局はどの様な種類の補償制度が実現可能で有効であると考えられるのかについてその意見を求めている。著作権局は、元の学習マテリアルを特定するのに必要な記録の保持とこれらの情報の著作権者及びその他の者に対する入手可能性に関する情報も求めている。

2つ目の事項について、著作権局は、生成AIを用いて創作された物に対する適切な著作権保護の範囲に関するコメントを求める。法律上アメリカ合衆国における著作権保護は人間の著作による作品に限られる事は明らかであると私たちは信じているが(原注:2023年8月18日のターラー対パールムッター事件コロンビア地裁判決(AI生成物に対する著作権局の登録拒絶を支持している)参照。)、人間の創作とAI生成コンテンツの間の線をどこにどの様に引くべきかという事について疑問は残っている。例えば、学習マテリアルの選択及び指示(「プロンプト」)の多くの繰り返しを通じて、生成AIシステムの人間の利用が技術に対して十分なコントロールを含み、人間の著作である出力をもたらす様な状況があるのか?この疑問の解決は将来の登録決定に影響する。著作権局は別途AI生成物を含む作品に関する登録指針の更新に取り組んでいるが(原注:例えば、著作権局は、申請人はその著作物における最小限度の量を超えるAI生成コンテンツを開示をするとしている指針をどの様に適用すべきかに関する質問を受け取っている。AI登録指針(「最小限度を超えるAI生成コンテンツは申請から明示的に覗かれるべきである」と説明している)参照。)、この意見募集は著作権保護適格性に関するより広い政策事項を調査するものである。

3つ目の事項について、著作権局は、著作権の責任の原則が生成AIシステムによって創作された物にどの様に適用され得るのかについて関心を持っている。(これらの質問の幾つかは現在生成AIシステムに関して既に提起された訴訟において裁判所に掛かるものとなっている。例えば、カリフォルニア北部地裁のJ.L.対アルファベット社事件;同地裁のカドレー対メタ・プラットフォーム社事件;同地裁のシルバーマン対オープンAI社事件;デラウェア地裁のゲッティ・イメージズ社対スタビリティAI社事件;カリフォルニア北部地裁のアンダーセン対スタビリティAI社事件;同地裁のドゥ対ギットハブ社事件参照。)例えば、出力が学習データセットの一部をなす著作物と実質的に同一となり、その利用が構成と認められない場合に、その指示により出力をもたらした利用者とシステムとデータセットの開発者の間で責任はどの様に配分されるべきだろうか?

最後に、私たちのヒアリング会合及びその他の呼び掛けの両方において、著作権局は、その声、肖像又はスタイルを真似る生成AIシステムの能力に対するアーティスト及び実演家の懸念の声を聞いた。これらの個人の属性は一般的に著作権法によって保護されるものではないが、これらの複製行為は、様々な州のパブリシティ権及び不正競争法に関わるものであり、様々な国際条約の義務に関係を持つものである(2019年4月のアメリカ著作権局報告書、著作者、帰属及び完全性:アメリカ合衆国における人格権の調査、https://www.copyright.gov/policy/moralrights/full-report.pdf(この様な利益が一般的にどの様に州のパブリシティ権法によって保護されるのかを議論している)参照)。

 さらに詳細な質問も作られており、関心のある方は是非元の募集要項全体も御覧いただければと思うが、上で抜き出した部分だけを見ても分かる様に、このアメリカ著作権局の意見募集は、AI学習のための著作物の利用、AI生成物の著作権保護適格性、AI生成物による著作権侵害と、AIと著作権の問題に関して考えられるほぼ全ての論点を網羅した包括的なものとなっている。

 この意見募集の〆切は10月18日であり(追加コメント期間は11月15日まで)、その取りまとめを受けてアメリカ政府又は議会でさらに検討が進められるのは来年以降になるのではないだろうか。

 今回のコロンビア地裁の判決も、意見募集ペーパーも特に何か新しい事を示すものではないが、どこの国であれAIと著作権の関係に関する論点は同じであって、これらはアメリカにおける法的議論の現在地点を示すものとして興味深いものである。

 意見募集の結果の取りまとめも注目に値するものとなるだろうが、現時点では、アメリカであっても、関係訴訟が多く提起されている中で何らかの立法的解決を図る事は難しく、判例による著作権法の合理的解釈を通じた解決が進んで行くのではないか、いずれの事件も最高裁まで行く可能性がある事を考えても、最終的に何らかの方向性が見えるまでにはまだかなりの時間が掛かるのではないかというのが私の予想である。AI技術の研究開発の中心であり、否応なく世界的に大きな影響を及ぼすだろう、アメリカの著作権動向については、ここでも折に触れ紹介して行きたいと考えている。

(2024年2月26日の追記:誤記を1箇所修正した(「構成s」→「構成」)。)

| | コメント (0)

2023年4月 9日 (日)

第476回:アメリカ著作権局のAI生成物の著作権登録を不可とする方針ペーパー

 この3月16日にアメリカ著作権局がAI生成物の著作権登録を不可とする方針を示す文書を公表した。

 この方針ペーパー(pdf)は、世界の主要な著作権当局が示したAI生成物の著作権法上の取り扱いを示した最初の文書であり、アメリカの主要な判例も含め良くまとまった非常に興味深いものと思うので、ここでその内容について取り上げたい。

 まず、かなり長くなるが、このペーパーから、AI生成物が著作権保護の対象となるかという事に関する実質的な判断に関する事を示した部分を以下に訳出する。(いつも通り、翻訳は拙訳。)

I. Background

...

One such recent development is the use of sophisticated artificial intelligence ("AI") technologies capable of producing expressive material. (note 5: The term "expressive material" is used here to refer to AI output that, if it had been created by a human, would fall within the subject matter of copyright as defined in section 102 of the Act. See id. at 102(a).) These technologies "train" on vast quantities of preexisting human authored works and use inferences from that training to generate new content. Some systems operate in response to a user's textual instruction, called a "prompt." (note 6: See Prompts, Midjourney, https://docs.midjourney.com/docs/prompts (noting for users of the artificial intelligence service Midjourney a prompt is "a short text phrase that the Midjourney [service] uses to produce an image"). To be clear, this policy statement is not limited to AI technologies that accept text "prompts" or to technologies permitting prompts of a particular length or complexity.) The resulting output may be textual, visual, or audio, and is determined by the AI based on its design and the material it has been trained on. These technologies, often described as "generative AI," raise questions about whether the material they produce is protected by copyright, whether works consisting of both human-authored and AI-generated material may be registered, and what information should be provided to the Office by applicants seeking to register them.

These are no longer hypothetical questions, as the Office is already receiving and examining applications for registration that claim copyright in AI-generated material. For example, in 2018 the Office received an application for a visual work that the applicant described as "autonomously created by a computer algorithm running on a machine." (note 7: U.S. Copyright Office Review Board, Decision Affirming Refusal of Registration of a Recent Entrance to Paradise at 2 (Feb. 14, 2022), https://www.copyright.gov/rulings-filings/review-board/docs/a-recent-entrance-to-paradise.pdf.) The application was denied because, based on the applicant's representations in the application, the examiner found that the work contained no human authorship. After a series of administrative appeals, the Office's Review Board issued a final determination affirming that the work could not be registered because it was made "without any creative contribution from a human actor." (note 8: Id. at 2-3. The Office's decision is currently being challenged in Thaler v. Perlmutter, Case No. 1:22-cv-01564 (D.D.C.).)

More recently, the Office reviewed a registration for a work containing human-authored elements combined with AI-generated images. In February 2023, the Office concluded that a graphic novel (note 9: On the application, the applicant described the work as a "comic book." See U.S. Copyright Office, Cancellation Decision re: Zarya of the Dawn (VAu001480196) at 2 (Feb. 21, 2023), https://www.copyright.gov/docs/zarya-of-the-dawn.pdf.) comprised of human authored text combined with images generated by the AI service Midjourney constituted a copyrightable work, but that the individual images themselves could not be protected by copyright. (note 10: Id.)

The Office has received other applications that have named AI technology as the author or co-author of the work or have included statements in the "Author Created" or "Note to Copyright Office" sections of the application indicating that the work was produced by or with the assistance of AI. Other applicants have not disclosed the inclusion of AI-generated material but have mentioned the names of AI technologies in the title of the work or the "acknowledgments" section of the deposit.

Based on these developments, the Office concludes that public guidance is needed on the registration of works containing AI-generated content. This statement of policy describes how the Office applies copyright law's human authorship requirement to applications to register such works and provides guidance to applicants.

The Office recognizes that AI-generated works implicate other copyright issues not addressed in this statement. It has launched an agency-wide initiative to delve into a wide range of these issues. Among other things, the Office intends to publish a notice of inquiry later this year seeking public input on additional legal and policy topics, including how the law should apply to the use of copyrighted works in AI training and the resulting treatment of outputs.

II. The Human Authorship Requirement

In the Office's view, it is well established that copyright can protect only material that is the product of human creativity. Most fundamentally, the term "author," which is used in both the Constitution and the Copyright Act, excludes non-humans. The Office's registration policies and regulations reflect statutory and judicial guidance on this issue.

In its leading case on authorship, the Supreme Court used language excluding non-humans in interpreting Congress's constitutional power to provide "authors" the exclusive right to their "writings." (note 11: U.S. Const. art. I, sec. 8, cl. 8 (Congress has the power "[t]o promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.").) In Burrow-Giles Lithographic Co. v. Sarony, a defendant accused of making unauthorized copies of a photograph argued that the expansion of copyright protection to photographs by Congress was unconstitutional because "a photograph is not a writing nor the production of an author" but is instead created by a camera. (note 12: 111 U.S. 53, 56 (1884) (explaining that the defendant had argued that photographs were merely "reproduction on paper of the exact features of some natural object or of some person").) The Court disagreed, holding that there was "no doubt" the Constitution's Copyright Clause permitted photographs to be subject to copyright, "so far as they are representatives of original intellectual conceptions of the author." (note 13: Id. at 58.) The Court defined an "author" as "he to whom anything owes its origin; originator; maker; one who completes a work of science or literature." (note 14: Id. at 57-58.) It repeatedly referred to such "authors" as human, describing authors as a class of "persons" (note 15: Id. at 56 (describing beneficiaries of the Constitution's Copyright Clause as "authors," who are one of "two classes" of "persons").) and a copyright as "the exclusive right of a man to the production of his own genius or intellect." (note 16: Id. at 58; see also id. at 60-61 (agreeing with an English decision describing an "author" as the "person" who was "the cause of the picture which is produced" and "the man" who creates or gives effect to the idea in the work).)

Federal appellate courts have reached a similar conclusion when interpreting the text of the Copyright Act, which provides copyright protection only for "works of authorship." (note 17: 17 U.S.C. 102(a).) The Ninth Circuit has held that a book containing words "authored by non-human spiritual beings" can only qualify for copyright protection if there is "human selection and arrangement of the revelations." (note 18: Urantia Found. v. Kristen Maaherra, 114 F.3d 955, 957-59 (9th Cir. 1997) (internal punctuation omitted) (holding that "some element of human creativity must have occurred in order for the Book to be copyrightable" because "it is not creations of divine beings that the copyright laws were intended to protect"). While the compilation of the book was entitled to copyright, the alleged "divine messages" were not. Id.) In another case, it held that a monkey cannot register a copyright in photos it captures with a camera because the Copyright Act refers to an author's "children," "widow," "grandchildren," and "widower," - terms that "all imply humanity and necessarily exclude animals." (note 19: Naruto v. Slater, 888 F.3d 418, 426 (9th Cir. 2018), decided on other grounds.)

Relying on these cases among others, the Office's existing registration guidance has long required that works be the product of human authorship. In the 1973 edition of the Office's Compendium of Copyright Office Practices, the Office warned that it would not register materials that did not "owe their origin to a human agent." (note 20: U.S. Copyright Office, Compendium of U.S. Copyright Office Practices sec. 2.8.3(I)(a)(1)(b) (1st ed. 1973), https://copyright.gov/history/comp/compendium-one.pdf (providing example of shapes formed by liquid petroleum); see also U.S. Copyright Office, Sixty-Eighth Annual Report of the Register of Copyrights for the Fiscal Year Ending June 30, 1965, at 5 (1966), https://www.copyright.gov/reports/annual/archive/ar-1965.pdf (noting that computer-generated works raise a "crucial question" of whether the work "is basically one of human authorship").) The second edition of the Compendium, published in 1984, explained that the "term 'authorship' implies that, for a work to be copyrightable, it must owe its origin to a human being." (note 21: U.S. Copyright Office, Compendium of U.S. Copyright Office Practices sec. 202.02(b) (2d ed. 1984), https://www.copyright.gov/history/comp/compendium-two.pdf (explaining that as a result, "[m]aterials produced solely by nature, by plants, or by animals are not copyrightable"). It went on to state that because "a work must be the product of human authorship," works "produced by mechanical processes or random selection without any contribution by a human author are not registrable." Id. at 503.03(a).) And in the current edition of the Compendium, the Office states that "to qualify as a work of 'authorship' a work must be created by a human being" and that it "will not register works produced by a machine or mere mechanical process that operates randomly or automatically without any creative input or intervention from a human author." (note 22: U.S. Copyright Office, Compendium of U.S. Copyright Office Practices sec. 313.2 (3d ed. 2021) ("Compendium (Third)").)

III. The Office's Application of the Human Authorship Requirement

As the agency overseeing the copyright registration system, the Office has extensive experience in evaluating works submitted for registration that contain human authorship combined with uncopyrightable material, including material generated by or with the assistance of technology. It begins by asking "whether the 'work' is basically one of human authorship, with the computer [or other device] merely being an assisting instrument, or whether the traditional elements of authorship in the work (literary, artistic, or musical expression or elements of selection, arrangement, etc.) were actually conceived and executed not by man but by a machine." (note 23: Id. (quoting U.S. Copyright Office, Sixty-Eighth Annual Report of the Register of Copyrights for the Fiscal Year Ending June 30, 1965, at 5 (1966)).) In the case of works containing AI-generated material, the Office will consider whether the AI contributions are the result of "mechanical reproduction" or instead of an author's "own original mental conception, to which [the author] gave visible form." (note 24: Sarony 111 U.S. at 60.) The answer will depend on the circumstances, particularly how the AI tool operates and how it was used to create the final work. (note 25: Many technologies are described or marketed as "artificial intelligence," but not all of them function the same way for purposes of copyright law. For that reason, this analysis will be fact specific.) This is necessarily a case-by-case inquiry.

If a work's traditional elements of authorship were produced by a machine, the work lacks human authorship and the Office will not register it. (note 26: This includes situations where an AI technology is developed such that it generates material autonomously without human involvement. See U.S. Copyright Office Review Board, Decision Affirming Refusal of Registration of a Recent Entrance to Paradise at 2-3 (Feb. 14, 2022), https://www.copyright.gov/rulings-filings/review-board/docs/a-recent-entrance-to-paradise.pdf (determining a work "autonomously created by artificial intelligence without any creative contribution from a human actor" was "ineligible for registration").) For example, when an AI technology receives solely a prompt (note 27: While some prompts may be sufficiently creative to be protected by copyright, that does not mean that material generated from a copyrightable prompt is itself copyrightable.) from a human and produces complex written, visual, or musical works in response, the "traditional elements of authorship" are determined and executed by the technology - not the human user. Based on the Office's understanding of the generative AI technologies currently available, users do not exercise ultimate creative control over how such systems interpret prompts and generate material. Instead, these prompts function more like instructions to a commissioned artist - they identify what the prompter wishes to have depicted, but the machine determines how those instructions are implemented in its output. (note 28: One image-generating AI product describes prompts as "influencing" the output but does not suggest the prompts dictate or control it. See Prompts, Midjourney, https://docs.midjourney.com/docs/prompts (explaining that short text prompts cause "each word [to have] a more powerful influence" and that images including in a prompt may "influence the style and content of the finished result") (emphasis added).) For example, if a user instructs a text-generating technology to "write a poem about copyright law in the style of William Shakespeare," she can expect the system to generate text that is recognizable as a poem, mentions copyright, and resembles Shakespeare's style. (note 29: AI technologies do not always operate precisely as instructed. For example, a text-generating tool prompted to provide factual information may provide inaccurate information. One AI service describes this as the AI "mak[ing] up facts or 'hallucinat[ing]' outputs." ChatGPT General FAQ, OpenAI, https://help.openai.com/en/articles/6783457-chatgpt-general-faq. See also James Romoser, No, Ruth Bader Ginsburg did not dissent in Obergefell - and other things ChatGPT gets wrong about the Supreme Court, SCOTUSblog (Jan. 26, 2023), https://www.scotusblog.com/2023/01/no-ruth-bader-ginsburg-did-not-dissent-in-obergefell-and-other-things-chatgpt-gets-wrong-about-the-supreme-court/.) But the technology will decide the rhyming pattern, the words in each line, and the structure of the text. (note 30: Some technologies allow users to provide iterative "feedback" by providing additional prompts to the machine. For example, the user may instruct the AI to revise the generated text to mention a topic or emphasize a particular point. While such instructions may give a user greater influence over the output, the AI technology is what determines how to implement those additional instructions.) When an AI technology determines the expressive elements of its output, the generated material is not the product of human authorship. (note 31: See id. at 61 (quoting British decision by Lord Justice Cotton describing an author as the person "who has actually formed the picture").) As a result, that material is not protected by copyright and must be disclaimed in a registration application. (note 32: See Compendium (Third) sec. 503.5 (a copyright registration "does not cover any unclaimable material that the work may contain," and applicants "should exclude that material from the claim").)

In other cases, however, a work containing AI-generated material will also contain sufficient human authorship to support a copyright claim. For example, a human may select or arrange AI-generated material in a sufficiently creative way that "the resulting work as a whole constitutes an original work of authorship." (note 33: 17 U.S.C. 101 (definition of "compilation"). In the case of a compilation including AI-generated material, the computer-generated material will not be protected outside of the compilation.) Or an artist may modify material originally generated by AI technology to such a degree that the modifications meet the standard for copyright protection. (note 34: See Compendium (Third) sec. 507.1 (identifying that where a new author modifies a preexisting work, the "new authorship ... may be registered, provided that it contains a sufficient amount of original authorship"); see also 17 U.S.C. 101 (defining "derivative work" to include works "based upon one or more preexisting works" where modifications to the work "which, as a whole, represent an original work of authorship").) In these cases, copyright will only protect the human-authored aspects of the work, which are "independent of" and do "not affect" the copyright status of the AI-generated material itself. (note 35: 17 U.S.C. 103(b).)

This policy does not mean that technological tools cannot be part of the creative process. Authors have long used such tools to create their works or to recast, transform, or adapt their expressive authorship. For example, a visual artist who uses Adobe Photoshop to edit an image remains the author of the modified image, (note 36: To the extent, however, that an artist uses the AI-powered features in Photoshop, the edits will be subject to the above analysis.) and a musical artist may use effects such as guitar pedals when creating a sound recording. In each case, what matters is the extent to which the human had creative control over the work's expression and "actually formed" the traditional elements of authorship. (note 37: Sarony, 111 U.S. at 61.)

...

Ⅰ.背景

(略)

この様な最近の発展には、表現物を作成する事が可能な洗練された人工知能(「AI」)技術の利用がある。(原注:「表現物」という語はここで、人間によって作り出されたものである場合に、アメリカ著作権法第102条において定義される著作権の対象になるものを指すために使われる。同第102条(a)参照。)この技術は、既存の人間によって大量の創作された著作物を「学習」し、その学習からの推論を用いて新しいコンテンツを生成する。あるシステムは、「プロンプト」と呼ばれる、利用者のテキストによる指示に対する応答で動作する。(原注:ミッドジャーニー、プロンプトの説明、https://docs.midjourney.com/docs/prompts
(人工知能サービスのミッドジャーニーの利用者に対し、プロンプトは「ミッドジャーニー[サービス]が画像を作成するのに使う短い文章のフレーズ」であると注意している)参照。明確化として、本方針文書は、文章「プロンプト」を受け入れるAI技術又は特定の長さ若しくは複雑さのプロンプトを許容する技術に限定されるものではない。)得られる出力は文章、視覚的なもの、音声であり得、それはその設計及び学習されたマテリアルに基づくAIによって決まる。「生成AI」と良く書かれるこの技術から、それが作成するマテリアルが著作権によって保護されるかどうか、そして、それを登録しようとする申請者によりアメリカ著作権局にどの様な情報が提供されるべきかという質問が発生している。

アメリカ著作権局は既にAI生成物に著作権を請求する登録申請を受け、審査しており、それらはもはや仮定の質問ではない。例えば、2018年に、著作権局は、申請者が「計算機の上で実行されるコンピューターアルゴリズムによって自動的に作り出されたものである」と書いた画像に対する申請を受けた。(原注:アメリカ著作権局再審査部、天国への最近の入国の登録の拒絶を肯定する決定(2022年2月14日)、https://www.copyright.gov/rulings-filings/review-board/docs/a-recent-entrance-to-paradise.pdf。)申請者の申請における記載に基づき、審査官が作品は人間による創作を含んでいないと判断したため、申請は拒絶された。一連の行政不服の後、著作権局の再審査部は、「人間の参加者からの創造的寄与に全くよらずに」作られたものであるから、その作品は登録され得ないという事を肯定する最終決定を出した。(原注:著作権局の決定に対し現在ターラー対パールムッター事件として訴訟が提起されている。)

さらに最近、アメリカ著作権局は、AIにより生成された画像と組み合わされた人間により創作された要素を含む作品に対する登録を審査した。2023年2月に、著作権局は、AIサービスであるミッドジャーニーによって生成された画像と組み合わされた人間により創作された文章を含むグラフィックノベル(原注:申請において、申請者は作品を「コミックブック」と書いている。アメリカ著作権局、黄昏のザリヤ事件取り消し決定(2023年2月21日)、https://www.copyright.gov/docs/zarya-of-the-dawn.pdf参照。)は、著作権の保護を受ける著作物を構成するが、個々の画像自体は著作権によって保護されないと結論づけた。(原注:同上。)

アメリカ著作権局は、AI技術を著作者又は共同著作者の名にあげるか、申請の「作り出した著者」又は「著作権局に対する注意」の項目に作品がAIの補助によるか又はそれとともに作成された事を示す記載を含む他の申請も受けた。他の申請者はAI生成物が含まれる事を開示しないが、提出物の作品の表題又は「謝辞」の項目にAI技術の名前を言及していた。

この展開に基づき、アメリカ著作権局は、AI生成コンテンツを含む作品の登録に関し、公衆ガイダンスが必要であると結論づける。この方針文書は、どの様に著作権局が著作権法の人間による創作の要件を申請に適用してその様な作品を登録するかを記載し、申請者にガイダンスを提供する。

アメリカ著作権局は、AI生成物について本文書で取り扱わない他の著作権問題がある事を認める。この問題を広い範囲で探究するため局全体の取り組みを開始している。とりわけ、著作権局は、どの様に法律がAI学習における著作物の利用及び出力の結果の取り扱いに適用されるべきかなど、その他の法的及び政策的事項に関する公衆のインプットを求める調査の通知を今年の後半に公表する事を考えている。

Ⅱ.人間による創作の要件

アメリカ著作権局の考えでは、著作権は人間の創造性の作成物であるマテリアルだけを保護できるという事は良く確立されている。最も根本的な事として、アメリカ憲法及び著作権法の両方において用いられている「著作者」という語は人間でないものを排除している。著作権局の登録ポリシー及び規則はこの事に関する立法及び司法のガイダンスを反映している。

人間による創作性に関するリーディングケースにおいて、最高裁は、「著作者」にその「著作」に排他的権利を与える、アメリカ議会の憲法権限の解釈において、人間でないものを排除する言葉を用いた。(原注:アメリカ憲法第1条第8節第8項(議会は「著作者及び発明者にそれぞれの著作及び発見に対して限定された期間排他的権利を保障する事により科学と有用な芸術の発展を促進する」権限を有する。)。)ビュロー・ジル・リトグラフ対サロニー事件において、写真の無許諾複製の作成について訴えられた被告は、「写真は著作者の著作でも作成物でもなく」カメラによって作り出されたものであるから、議会による写真への著作権保護の拡大は違憲であると主張した。(原注:1884年3月17日のアメリカ最高裁判決(被告は、写真は単に「ある自然物かある人間の正確な特徴の紙の上の再生」であると主張したと説明している)。)最高裁はこれに同意せず、「それが著作者の独創的、知的着想を表現したものである限り」、憲法の著作権条項は写真が著作権の対象である事を認めている事に「疑いはない」と判示した。(原注:同上。)裁判所は、「著作者」を「何であれその起源を負う者;又は創造者;製作者;科学又は文学の作品を完成した者」と定義した。(原注:同上。)最高裁は、著作者は「人」の一種であると述べ、繰り返し、その様な「著作者」を人間として(原注:同上(憲法の著作権条項の受益者は「人」の2つの種類の内の1つである「著作者」と述べている)。)、著作権を「その才能又は知性の作成物について人に与えられる排他的権利」として言及している。(原注:同上(「著作者」は「作成された絵の起因」である「人」であり、作り出したかその着想をその作品に結実させた「人」であると述べたイギリスの判決に同意している)。)

連邦控訴裁判所は、「創作による作品」に対してのみ著作権保護を与える、著作権法の条文解釈において同様の結論に達している。(原注:アメリカ著作権法第102条(a)。)第9巡回控訴裁は、「人間でない霊的な存在によって創作された」言葉を含む本は、「啓示の人間による選択及び配列」がある場合にのみ、著作権保護を受けられると判示した。(原注:1997年6月10日のウランティア協会対クリステン・マアヘラ事件第9巡回控訴裁判決(「著作権法が保護する事を意図しているのは聖なる存在の創造ではない」のであるから、「本が著作権の保護を受けるためには何がしかの人間の創造性の要素が生じていなければならない」と判示している)。本の編集が著作権の保護を受けるのに対し、主張されている「聖なる通知」はそうではない。同上。)他の事件において、同控訴裁は、著作権法は著作者の「子供」、「寡婦」、「孫」及び「寡夫」-「全て人間を示唆し、必然的に動物を排除する」用語-について言及している事から、猿がカメラによって撮った写真において猿は著作権を登録できないと判示した。(原注:2018年4月23日のナルト対スレーター事件第9巡回控訴裁判決、ただし、これは他の理由により決定がなされた。)

とりわけこれらのケースに依拠し、アメリカ著作権局の既存の登録ガイダンスは、長く、作品は人間の創作による作成物である事を求めて来た。1973年版の著作権局の著作権局実務便覧において、著作権局は、その起源を人間の主体に負わないマテリアルを登録しないと注意した。(原注:アメリカ著作権局、著作権局実務便覧2.8.3(Ⅰ)(a)(1)(b)(1973年第1版)、https://copyright.gov/history/comp/compendium-one.pdf);及びアメリカ著作権局、1965年6月30日までの年度の第68著作権登録年次報告書(1966年)、https://www.copyright.gov/reports/annual/archive/ar-1965.pdf(コンピュータ生成作品が、その作品が「基本的に人の創作によるものである」かどうかについての「重大な問題」を生じさせていると記載している)。)1984年に公表された便覧の第2版は、「『創作性』という語は、著作権の保護を受ける作品はその起源を人間に負わなければならない事を示唆している」と説明している。(原注:アメリカ著作権局、著作権局実務便覧202.02(b)(1984年第2版)、https://www.copyright.gov/history/comp/compendium-two.pdf(結果として、「自然、植物又は動物のみによって作成されたマテリアルは著作権の保護を受けられない」と説明している)。また、これは「作品は人の創作性の作成物でなければならない」事から、人の著作者による何らの寄与なく機械的なプロセス又はランダムの選択によって作成された作品は登録され得ない」と記載している。同上503.03(a)。)そして、便覧の現在版において、著作権局は、『創作性』による作品と評価されるために作品は人間によって作り出されたものでなければならない」のであり、機械によってか、何ら人間の著作者による創造的入力又は介入なくランダムに又は自動的に動作する機械的プロセスによって作成された作品を登録しないと記載している。(原注:アメリカ著作権局、著作権局実務便覧313.2(2021年第3版))

Ⅲ.人間による創作の要件のアメリカ著作権局における適用

著作権登録制度を監督する主体として、アメリカ著作権局は技術によってかその補助を受けて作り出されたマテリアルなど、著作権の保護を受けられないマテリアルと組み合わされた、人間の創作性を含む、登録のために提出された作品の評価に関して幅広い経験を有している。著作権局はまず、「『作品』が基本的に人間の創作によるものであり、コンピューター[又は他の機器」は単に補助的道具であるのか、又は、作品における創作性の伝統的要素(文学的、芸術的又は音楽的表現又は選択、配列等の要素)は機械によってではなく実際に人によって思いつかれたか実行されたか」を検討する。(原注:同上(アメリカ著作権局、1965年6月30日までの年度の第68著作権登録年次報告書(1966年)を引用している)。)AI生成マテリアルを含む作品の場合に、著作権局は、AIの寄与が「機械的再生」によるものであるか、それとも著作者が「視認可能な形式を与えた自身の独創的、精神的着想」によるものであるのかを検討する。(原注:サロニー事件判決。)答えは状況により、特にどの様にAIツールが動作し、どの様にそれが最終的な作品を作り出すのに使われたかによる。(原注:多くの技術が「人工知能」として記述され、市場で提供されているが、その全てが著作権法の目的において同じ様に機能する訳ではない。この理由から、この分析は事実による。)これは必然的にケースバイケースの調査となる。

作品の創作性の伝統的要素が機械によって作成された場合、作品は人間による創作の要件を欠き、アメリカ著作権局はそれを登録しない。(原注:これは、AI技術が人間の関与なく自動的にマテリアルを生成する様に開発された状況を含む。アメリカ著作権局再審査部、天国への最近の入国の登録の拒絶を肯定する決定(2022年2月14日)、https://www.copyright.gov/rulings-filings/review-board/docs/a-recent-entrance-to-paradise.pdf(「何ら人間の主体からの創造的寄与なく人工知能によって自動的に作り出された」作品は「登録を認められない」と決定している)参照。)例えば、AI技術が単に人間からのプロンプト(あるプロンプトに十分創造性があり、著作権によって保護を受けられる事はあり得るであろうが、その事は著作権の保護を受けられるプロンプトによって生成されたマテリアルがそれ自体著作権の保護を受けられる事を意味しない。)を受け取り、応答として複雑な文章の、視覚的又は音楽の作品を作成するとき、「創作性の伝統的要素」は-人間の利用者ではなく-技術によって決定され、実行されている。現在利用可能な生成AI技術に対する著作権局の理解に基づくと、利用者は、どの様にその様なシステムがプロンプトを解釈し、マテリアルを生成するかについての最終的な創造的コントロールを行っていない。これらのプロンプトはどちらかと言えば委託されたアーティストに対する指示の様に機能する-それらはプロンプトを書いた者が描いて欲しいものを特定するが、機械がどの様にその指示を出力で実施するかを決定している。(原注:ある画像生成AI製品は、「プロンプト」は出力に影響を与えるが、プロンプトがそれを支配したり、コントロールする事を示すものではないと述べている。(原注:ミッドジャーニー、プロンプトの説明、https://docs.midjourney.com/docs/prompts (短い文章のプロンプトは「それぞれの語が強力な影響を」もたらすものであり、あるプロンプトに含まれるイメージは「最終結果のスタイルと内容に影響し得る」と説明している)参照。)例えば、もしある利用者が「ウィリアム・シェイクスピアの文体で著作権法に関する詩を書く」事を文章生成技術に指示したら、システムが、詩と認識可能で、著作権に言及し、シェイクスピアの文体に似ていると認識できる文章を生成する事を期待できる。(原注:AI技術は必ずしも指示した通り正確に動作しない。例えば、事実である情報を提供するようプロンプト入力された文章生成ツールが不正確な情報を提供する事がある。あるAIサービスはこれを「事実を作り出すか出力において『錯覚』するAIと記述している。ChatGPT一般質問、OpenAI、https://help.openai.com/en/articles/6783457-chatgpt-general-faq。また、ジェイムズ・ロモーザー、否、ルース・ベイダー・ギンスブルクはオーバーゲフェル事件において反対意見を述べているのではない-最高裁についてその他ChatGPTが間違っている事、SCOTUSblog(2023年1月26日)、https://www.scotusblog.com/2023/01/no-ruth-bader-ginsburg-did-not-dissent-in-obergefell-and-other-things-chatgpt-gets-wrong-about-the-supreme-court/。)しかし、この技術は韻律のパターン、それぞれの行の語及び文章の構造を決定するであろう。(原注:技術には機械に対して追加のプロンプトを与える事によりインタラクティブな「フィードバック」を提供する事を利用者に可能としているものもある。例えば、利用者は生成された文章をあるトピックを言及する様に又は特定のポイントを強調する様に改めるようAIに指示する事ができる。その様な指示は出力に対してより大きな影響を与え得るものであるが、どの様にこれらの追加の指示を実施するかを決定しているのはAI技術である。)あるAI技術が出力の表現要素を決定するとき、生成されたマテリアルは人間による創作の作成物ではない。(原注:同上(著作者は「絵を実際に形作った」人であると述べるコットン判事によるイギリスの判決を引用している)。)つまり、そのマテリアルは著作権によって保護されず、登録申請書において覗かれなければならない。(原注:便覧(第3版)503.5(著作権登録は「その作品が含み得る請求不可能なマテリアルをカバーせず」、申請は「そのマテリアルを請求から除くべきである」)参照。)

しかしながら、他の場合において、AI生成物を含む作品は著作権の請求を支持するのに十分な人間による創作を含むであろう。例えば、「得られた作品が全体として創作による独創的な作品を構成する」のに十分創造的なやり方で人がAI生成物を選又は配列する事があり得る。(原注:アメリカ著作権法第101条(「編集」の定義)。AI生成物を含む編集の場合、コンピュータ生成物は編集の外で保護されないであろう。)あるアーティストが、修正が著作権保護の基準に合致する程度まで、元々AI技術によって生成されたマテリアルを修正する事もあり得る。(原注:便覧(第3版)507.1(新しい著作者が既存の作品を修正する場合、「新しい創作は…独創的創作性の十分な量を含む限りにおいて、登録され得る」と記載している)参照;アメリカ著作権法第101条(1つ以上の既存の作品に対する修正が「全体として創作による独創的な作品を表す」場合の、既存の作品に基づく作品を含む「派生著作物」を定義している)も参照。)これらの場合において、著作権は作品の人間により創作された側面のみを保護し、この事はAI生成物自体の著作権状態とは関係なく、また、これにに影響を与える事もないであろう。(原注:アメリカ著作権法第103条(b)。)

この方針は技術的ツールが創造的プロセスの部分であり得ない事を意味するものではない。著作者は長くその様なツールを用いてその作品を作り出すか、その表現による創作を作り直すか、変形するか、適応させて来たのである。例えば、画像を編集するのにアドビのフォトショップを用いる視覚アーティストは修正された画像の著作者であり続け(原注:しかしながら、アーティストがフォトショップでAIによる助力を受けた特徴を使う程度に応じて、編集は上記の分析の対象となるであろう。)、そして、音楽アーティストは音声レコーディングを作り出すときにギターペダル様なエフェクトを使う事ができる。それぞれの場合において、問題となる事はどの程度人間が作品の表現に対して創造的コントロールを有し、創作の伝統的な要素を「実際に形作った」のかである。(原注:サロニー事件判決。)

(以下略)

 この方針ペーパーの背景に、アメリカでは著作権侵害訴訟を提起するにあたり著作権局での登録が必須とされているため、著作権の登録が強く求められ、他の国と比べて大量の著作権登録申請がされており、その審査が行われているという事があるのは注意しておく必要がある。(この事は、2019年3月4日のフォース・エステート・パブリック・ベネフィット社対Wall-Street.com事件最高裁判決(pdf)で最高裁レベルで確認されている。アメリカ著作権法のこの解釈・運用自体、無登録主義を前提とするべき著作権法のあり方として問題であると思っているが、その事はここではひとまずおく。)

 その事情から登録基準を示す必要に迫られたアメリカ著作権局は、このペーパーで、プロンプト入力のみにより機械が自動的に生成した文章や画像などの生成物(AI生成物)について著作権登録を認めず、結果としてAI生成物は著作権によって保護されないとする方針をはっきりと示しているのである。

 上の抜粋とその翻訳で辿ってもらってもいいと思うが、このペーパー中で引用されている判例等を時系列順に極簡単な概要ともに並べておくと、以下の様になる。

 今までの所、アメリカの判例と著作権局の決定が示しているのは、著作権の保護を受けるためには最終的な表現において人の創作的関与を必要とするという事であるが、これは妥当なものと私も考える。

  今後の技術の発展、人が最終的な表現を創作的にコントロールできるかという事にもよって、どこまで技術の利用者である人の著作権が認められるかについてさらに事例が蓄積されて行く事だろうが、人ではないものが自然に、機械的又は自動的に生成した表現が著作権保護の対象となり得ないという事は、著作権法の様な創作保護法における最も基本的な原理の1つであって、これは技術的に人の創作物と見紛う文章や画像が機械的に生成され得る様になっているという事によって左右されるべきではないだろう。

 また、著作権との関係では、AI生成物が著作権保護の対象となるかどうかという点とは別の重要な論点は、上のペーパーでも今後パブリックコメントを行う予定と書かれている、機械学習への著作物の利用と結果として得られる生成物の提供が著作権侵害となるかどうかという点である。アメリカ著作権局が予定するパブリックコメントも重要だろうが、この点についてはアメリカでも別の訴訟になっており、AI技術に対して大きな影響を及ぼすだろうその動向には私も大いに注目している。

 次は少し間が空くかもしれないが、現時点でのAI生成物と日本の著作権法との関係について私なりのまとめを書きたいと思っている。

(2023年4月10日夜の追記:上の抜粋と翻訳部分でも辿りやすいようリンクを追加するとともに、幾つか誤記を直し、少し文章を整えた。)

| | コメント (0)

2021年5月23日 (日)

第440回:JavaのAPI利用に関するアメリカ最高裁フェアユース判決

 前々回取り上げた特許法等改正案が、5月13日の参議院経済産業委員会で全会一致で可決され、14日の参議院本会議で可決・成立した。前回取り上げた著作権法改正案も、14日の衆議院文部科学委員会で全会一致で可決され、18日の衆議院本会議で可決されて参議院に送られ、20日の参議院文教科学委員会で趣旨説明がされた所なので、じきに委員会で審議され、本会議で可決・成立となるだろう。(それぞれ、衆議院のインターネット審議中継、参議院のインターネット審議中継参照。)

 その間に、4月5日に、アメリカの最高裁判所が、JavaのAPI利用はフェアユースであるとする判決を出した。これは、技術的な事も絡み、アメリカにおけるフェアユースの考え方とその広がりを見る上で非常に興味深いものと思うので、今回は、この判決について紹介する。

 まず、その判決(pdf)の最初の概要部分を以下に訳出する。

Oracle America, Inc., owns a copyright in Java SE, a computer platform that uses the popular Java computer programming language. In 2005, Google acquired Android and sought to build a new software platform for mobile devices. To allow the millions of programmers familiar with the Java programming language to work with its new Android platform, Google copied roughly 11,500 lines of code from the Java SE program. The copied lines are part of a tool called an Application Programming Interface (API). An API allows programmers to call upon prewritten computing tasks for use in their own programs. Over the course of protracted litigation, the lower courts have considered (1) whether Java SE's owner could copyright the copied lines from the API, and (2) if so, whether Google's copying constituted a permissible "fair use" of that material freeing Google from copyright liability. In the proceedings below, the Federal Circuit held that the copied lines are copyrightable. After a jury then found for Google on fair use, the Federal Circuit reversed, concluding that Google's copying was not a fair use as a matter of law. Prior to remand for a trial on damages, the Court agreed to review the Federal Circuit's determinations as to both copyrightability and fair use.

Held: Google's copying of the Java SE API, which included only those lines of code that were needed to allow programmers to put their accrued talents to work in a new and transformative program, was a fair use of that material as a matter of law. Pp. 11-36.

(a) Copyright and patents, the Constitution says, serve to "promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." Art. I, §8, cl. 8. Copyright encourages the production of works that others might cheaply reproduce by granting the author an exclusive right to produce the work for a period of time. Because such exclusivity may trigger negative consequences, Congress and the courts have limited the scope of copyright protection to ensure that a copyright holder's monopoly does not harm the public interest.

This case implicates two of the limits in the current Copyright Act. First, the Act provides that copyright protection cannot extend to "any idea, procedure, process, system, method of operation, concept, principle, or discovery . . . ." 17 U. S. C. §102(b). Second, the Act provides that a copyright holder may not prevent another person from making a "fair use" of a copyrighted work. §107. Google's petition asks the Court to apply both provisions to the copying at issue here. To decide no more than is necessary to resolve this case, the Court assumes for argument's sake that the copied lines can be copyrighted, and focuseson whether Google's use of those lines was a "fair use." Pp. 11-15.

(b) The doctrine of "fair use" is flexible and takes account of changes in technology. Computer programs differ to some extent from many other copyrightable works because computer programs always serve a functional purpose. Because of these differences, fair use has an important role to play for computer programs by providing a context-based check that keeps the copyright monopoly afforded to computer programs within its lawful bounds. Pp. 15-18.

(c) The fair use question is a mixed question of fact and law. Reviewing courts should appropriately defer to the jury's findings of underlying facts, but the ultimate question whether those facts amount to a fair use is a legal question for judges to decide de novo. This approach does not violate the Seventh Amendment's prohibition on courts reexamining facts tried by a jury, because the ultimate question here is one of law, not fact. The "right of trial by jury" does not include the right to have a jury resolve a fair use defense. Pp. 18-21.

(d) To determine whether Google's limited copying of the API here constitutes fair use, the Court examines the four guiding factors set forth in the Copyright Act's fair use provision: the purpose and character of the use; the nature of the copyrighted work; the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and the effect of the use upon the potential market for or value of the copyrighted work. §107. The Court has recognized that some factors may prove more important in some contexts than in others. Campbell v. Acuff-Rose Music, Inc., 510 U. S. 569, 577. Pp. 21-35.

(1) The nature of the work at issue favors fair use. The copied lines of code are part of a "user interface" that provides a way for programmers to access prewritten computer code through the use of simple commands. As a result, this code is different from many other types of code, such as the code that actually instructs the computer to execute a task. As part of an interface, the copied lines are inherently bound together with uncopyrightable ideas (the overall organization of the API) and the creation of new creative expression (the code independently written by Google). Unlike many other computer programs, the value of the copied lines is in significant part derived from the investment of users (here computer programmers) who have learned the API's system. Given these differences, application of fair use here is unlikely to undermine the general copyright protection that Congress provided for computer programs. Pp. 21-24.

(2) The inquiry into the "the purpose and character" of the use turns in large measure on whether the copying at issue was "transformative," i.e., whether it "adds something new, with a further purpose or different character." Campbell, 510 U. S., at 579. Google's limited copying of the API is a transformative use. Google copied only what was needed to allow programmers to work in a different computing environment without discarding a portion of a familiar programming language. Google's purpose was to create a different task-related system for a different computing environment (smartphones) and to create a platform - the Android platform - that would help achieve and popularize that objective. The record demonstrates numerous ways in which reimplementing an interface can further the development of computer programs. Google's purpose was therefore consistent with that creative progress that is the basic constitutional objective of copyright itself. Pp. 24-28.

(3) Google copied approximately 11,500 lines of declaring code from the API, which amounts to virtually all the declaring code needed to call up hundreds of different tasks. Those 11,500 lines, however, are only 0.4 percent of the entire API at issue, which consists of 2.86 million total lines. In considering "the amount and substantiality of the portion used" in this case, the 11,500 lines of code should be viewed as one small part of the considerably greater whole. As part of an interface, the copied lines of code are inextricably bound to other lines of code that are accessed by programmers. Google copied these lines not because of their creativity or beauty but because they would allow programmers to bring their skills to a new smartphone computing environment. The "substantiality" factor will generally weigh in favor of fair use where, as here, the amount of copying was tethered to a valid, and transformative, purpose. Pp. 28-30.

(4) The fourth statutory factor focuses upon the "effect" of the copying in the "market for or value of the copyrighted work." §107(4). Here the record showed that Google's new smartphone platform is not a market substitute for Java SE. The record also showed that Java SE's copyright holder would benefit from the reimplementation of its interface into a different market. Finally, enforcing the copyright on these facts risks causing creativity-related harms to the public. When taken together, these considerations demonstrate that the fourth factor - market effects - also weighs in favor of fair use. Pp. 30-35.

(e) The fact that computer programs are primarily functional makes it difficult to apply traditional copyright concepts in that technological world. Applying the principles of the Court's precedents and Congress' codification of the fair use doctrine to the distinct copyrighted work here, the Court concludes that Google's copying of the API to reimplement a user interface, taking only what was needed to allow users to put their accrued talents to work in a new and transformative program, constituted a fair use of that material as a matter of law. In reaching this result, the Court does not overturn or modify its earlier cases involving fair use. Pp. 35-36.

886 F. 3d 1179, reversed and remanded.

オラクルアメリカ社は、良く使われるJavaコンピュータプログラミング言語を利用するコンピュータプラットフォームであるJavaSEの著作権者である。2005年に、グーグルはアンドロイドを取得し、モバイル機器のための新たなソフトウェアプラットフォームを構築しようとした。Javaプログラミング言語に馴れ親しんだ数百万のプログラマーが新しいアンドロイドプラットフォームで働ける様にするため、グーグルはおよそ11500行のコードをJavaSEプラットフォームからコピーした。コピーされた行はアプリケーション・プログラミング・インターフェース(API)と呼ばれるツールの部分である。APIはプログラマーにその自身のプログラムにおける利用のために前もって書かれている計算タスクを呼び出す事を可能とするものである。長期に渡る訴訟経過の中で、下級審は、(1)JavaSEの権利者はAPIのコピーされた行について著作権を持ち得るか、(2)もしそうであるなら、グーグルのコピー行為はこのマテリアルの許される「フェアユース(公正利用)」となり、グーグルを著作権侵害責任から免除するかについて判断して来た。下の審理において、連邦裁判所は、コピーされた行は著作権保護の対象であると判示した。そして、陪審員がグーグルについてフェアユースとした後、連邦裁判所はそれを覆し、法律事項として、グーグルのコピー行為はフェアユースではないと結論づけた。賠償の審理のために差し戻すより先に、当裁判所は連邦裁判所の著作権保護対象性とフェアユースの両方に関する判断について再審理する事に同意した。

判示:プログラマーがその獲得した才能により新しく変形的なプログラムについて働く事を可能とするのに必要なものであるコード行のみを含む、グーグルのJavaSE APIのコピー行為は、法律事項として、そのマテリアルのフェアユースである。第11~36ページ。

(a)憲法は、著作権及び特許は「限定された期間において著作者と発明者にそれぞれの著作と発見について排他的権利を保証する事により科学と有用な芸術の進歩を促進するためのものである」と言っている(第1条第8節第8項)。著作物は他人が安価に複製可能なものであり得るが、著作権はある期間著作物を製造する排他的権利を著作者に付与する事によりその製造を促すものである。この様な排他性はネガティブな結果を引き起こし得るものであるから、議会と裁判所は、著作権者の独占が公益を害さない事を保証するために著作権保護の範囲を限定して来た。

本ケースにおいては、現行の著作権法における2つの限定が関係している。第1に、この法は、著作権の保護は「いかなるアイデア、手続き、プロセス、システム、実行方法、概念、原理、又は発見・・・」には及ばないと規定している(アメリカ合衆国法第17編(著作権法)第102条(b))。第2に、この法は、著作権者は他人が著作物について「フェアユース」をする事を止められないと規定している(第107条)。グーグルの訴えは、これらの規定を問題のコピー行為に適用する事を裁判所に求めるものである。本ケースを解決するのに必要以上の事を判断しないよう、当裁判所は、議論のために、コピーされた行は著作保護の対象であり得ると仮定し、グーグルのこれらの行の利用がフェアユースであるかに集中する。第11~15ページ。

(b)「フェアユース」の法理は柔軟であり、技術の変化を考慮に入れるものである。コンピュータプログラムは常に機能的な目的に使われるものであるから、コンピュータプログラムはある程度他の多くの著作物とは異なるものである。この様な違いのために、フェアユースは、コンピュータプログラムに与えられる独占をその適法な範囲に留める、文脈に基づくチェックを提供する事により、コンピュータプログラムに対して重要な役割を果たす。第15~18ページ。

(c)フェアユースの問題は事実と法律の混合問題である。再審査法廷は、陪審員による前提となる事実認定に適切に従うべきであるが、その事実からフェアユースとなるかの最終的な問題は法的な問題であり、判事が再審理する事ができる。ここでの最終的な問題は法律の問題であり、事実の問題ではないから、このアプローチは、修正第7条の陪審員による判断された事実の法廷による再審査の禁止に反するものではない。「陪審員による審理の権利」は、フェアユースの抗弁を陪審員が解決する権利を含むものではない。第18~21ページ。

(d)グーグルの限定的なAPIのコピー行為がフェアユースとなるかについて判断するため、当裁判所は、著作権法のフェアユース規定に定められた4つの主要要素を検討する:利用の目的及び性質;著作物の性質;著作物全体との関係で利用された部分の量及び本質性;並びに著作物の潜在的市場又は価値に対するその利用の影響(第107条)。当裁判所は、幾つかの要素が、ある文脈において他より重要なものと考えられる事を認めて来た(キャンベル事件判決)。第21~35ページ。

(1)問題となる著作物の性質はフェアユースに有利である。コピーされたコード行は、プログラマーに単純なコマンドの利用によって前もって書かれているコンピューターコードにアクセスする方法を提供する「ユーザインターフェース」の部分である。つまり、このコードは、コンピュータによるタスクの実行を実際に指示するコードの様な他の多くの種類のコードとは異なる。インターフェース部分として、コピーされた行は著作権保護対象を有しないアイデア(APIの全体的な構成)及び新たな創作的表現(グーグルによって独立に書かれたコード)と本質的に結びついている。他の多くのコンピュータプログラムと違い、コピーされた行の価値はその大部分がAPIシステムを学んだ利用者(コンピュータプログラマー)の投資に由来するものである。この様な違いがある事から、ここでのフェアユースの適用は議会がコンピュータプログラムについて規定した一般的な著作権保護を危うくするものとはならないであろう。第21~24ページ。

(2)利用の「目的及び性質」に関する調査は、多く、そのコピー行為が「変形的」であるか、すなわち、それが「さらなる他の目的又は異なる性質とともに新しい何かを追加する」かに帰着する(キャンベル事件判決)。グーグルのAPIの限定的なコピー行為は変形利用である。グーグルは、プログラマーが異なるコンピューティング環境において慣れ親しんだプログラミング言語の一部分を捨てずに働く事を可能にするのに必要なだけのものをコピーした。グーグルの目的は異なるコンピューティング環境(スマートフォン)のために異なるタスク関連システムを作り出し、その目標を達成して広める事に役立つプラットフォーム-アンドロイドプラットフォーム-を作り出す事にある。記録は、インターフェースの再実装によりさらなるコンピュータプログラムの開発が可能となる数多くのやり方を示している。したがって、グーグルの目的は著作権自体の憲法上の基本的な目的である創作的進歩と合致する。第24~28ページ。

(3)グーグルは、APIから、仮想的に何百かの異なるタスクを呼び出すのに必要な全ての宣言コードに及ぶ約11500行の宣言コードをコピーした。これらの11500行は、しかしながら、全体で286万行からなる、問題となるAPI全体の0.4%に過ぎない。本ケースにおける「利用された部分の量及び本質性」を考慮するにあたり、11500コード行は遥かに大きな全体の1小部分と見られるべきである。インターフェース部分として、コピーされたコード行は、プログラマーによってアクセスされる他のコード行と分離不能な形で結びついている。グーグルがこれらの行をコピーしたのは、その創作性又は美しさのためではなくそれによりプログラマーがその技能を新しいスマートフォンコンピューティング環境に持ち込む事を可能とするためである。「本質性」の要素は、ここでの様に、コピー行為の量が妥当なかつ変形の目的に結びついている場合には、一般的にフェアユースに有利に働く。第28~30ページ。

(4)第4の法定要素は、「著作物の潜在的市場又は価値」におけるコピー行為の「影響」に焦点をあてている(第107条第4項)。ここで、記録は、グーグルの新しいスマートフォンプラットフォームはJavaSEの代替市場ではない事を示している。記録は、また、JavaSEの著作権者はそのインターフェースの他の市場への再実装から利益を受けるであろう事も示している。最後に、この事実に基づき、著作権をエンフォースする事は公衆に対し創造性に関して害をもたらす恐れがある。これらの事を考え合わせると、これらの考察は、第4の要素-市場への影響-もフェアユースに有利に働く事を示している。第30~35ページ。

(e)コンピュータプログラムが主として機能的なものであるという事実は伝統的な著作権の概念をこの技術的な世界に適用する事を困難にしている。本ケースにおける独特な著作物に当裁判所の判例と議会の法制化によるフェアユースの法理を適用する事により、当裁判所は、プログラマーがその獲得した才能により新しく変形的なプログラムについて働く事を可能とするのに必要なもののみを取る、ユーザインターフェースを再実装するためのグーグルのAPIのコピー行為は、法律事項として、そのマテリアルのフェアユースとなると結論づける。この結論に至るにあたり、当裁判所はそのフェアユースに関する先例を覆したり、修正したりはしていない。第35~36ページ。

連邦裁判決は破棄され、差し戻される。

 今回のアメリカ最高裁判決のポイントは上の概要部分に良くまとめられているが、ここでコピーされたコード行と書かれているJavaのAPIは、判決の第7ページで以下の様に書かれている様に、Java言語プラットフォームの実行コードそのもの(プログラムと言う時に通常想定されるもの)ではなく、プログラマーがプログラムを書く際に利用する、宣言コードのみである事に注意が必要である。(以下で参考として一緒に付録Bのダイヤグラム表も引用しておく。)

Now let us consider the example that the District Court used to explain the precise technology here. Id., at 980-981. A programmer wishes, as part of her program, to determine which of two integers is the larger. To do so in the Java language, she will first write java.lang. Those words (which we have put in bold type) refer to the "package" (or by analogy to the file cabinet). She will then write Math. That word refers to the "class" (or by analogy to the drawer). She will then write max. That word refers to the "method" (or by analogy to the recipe). She will then make two parentheses ( ). And, in between the parentheses she will put two integers, say 4 and 6, that she wishes to compare. The whole expression - the method call - will look like this: "java.lang.Math.max(4, 6)." The use of this expression will, by means of the API, call up a task-implementing program that will determine the higher number.

In writing this program, the programmer will use the very symbols we have placed in bold in the precise order we have placed them. But the symbols by themselves do nothing. She must also use software that connects the symbols to the equivalent of file cabinets, drawers, and files. The API is that software. It includes both the declaring code that links each part of the method call to the particular task-implementing program, and the implementing code that actually carries it out. (For an illustration of this technology, see Appendix B, infra.)

Now we can return to the copying at issue in this case. Google did not copy the task-implementing programs, or implementing code, from the Sun Java API. It wrote its own task-implementing programs, such as those that would determine which of two integers is the greater or carry out any other desired (normally far more complex) task. This implementing code constitutes the vast majority of both the Sun Java API and the API that Google created for Android. App. 212. For most of the packages in its new API, Google also wrote its own declaring code. For 37 packages, however, Google copied the declaring code from the Sun Java API. Id., at 106-107. As just explained, that means that, for those 37 packages, Google necessarily copied both the names given to particular tasks and the grouping of those tasks into classes and packages.

ここで、地方裁判所が正確な技術について説明するのに使った例を考えよう。プログラマーが、そのプログラムの部分として、2つの整数のどちらが大きいかを決定したいとする。Java言語においてそうするために、プログラマーは、まず、java.langと書く。これらの語(私たちが太字にしたもの)は「パッケージ」(又は例えて言うならファイルキャビネットに相当する)を参照するものである。そして、プログラマーはMathと書く。この語は「クラス」(又は例えて言うなら引き出しに相当する)を参照するものである。そして、プログラマーは、maxと書く。この語は「メソッド」(又は例えて言うならレシピに相当する)。そして、プログラマーは2つの括弧( )を書く。そして、括弧の間にプログラマーは比較したい2つの整数、例えば4と6、を入れる。全体の表現-メソッドコール-は次の様になる:「java.lang.Math.max(4, 6)」。この表現の利用は、APIの手段によって、より大きな数を決定するタスク実行プログラムを呼び出す。

このプログラムを書くにあたり、プログラマーは、私たちが太字にした記号そのものを私たちが置いた順序通りに正確に利用する。しかし、記号はそれ自体では何もしない。プログラマーは、その記号を、ファイルキャビネット、引き出し及びファイルの等価物に繋げるソフトウェアも使わなくてはならない。APIがそのソフトウェアである。それは、メソッドコールのそれぞれの部分を個別のタスク実行プログラムに結びつける宣言コード及びそれを実際に実行する実行コードの両方を含んでいる。

ここで、私たちは、本ケースにおいて問題となるコピー行為に戻る。グーグルはサンのJavaAPIからタスク実行プログラムや実行コードをコピーしていない。グーグルは、2つの整数の内どちらが大きいかを決めるものや他の求める(通常はより遥かに複雑な)タスクを実行するものなどについて、自身のタスク実行プログラムを書いた。この実行コードはサンのJavaAPIとグーグルがアンドロイドのために作り出したAPIの大部分を構成している(地裁判決付録第212ページ)。その新しいAPIのパッケージのほとんどについて、グーグルは自身の宣言コードを書く事もしている。しかしながら、37のパッケージについて、グーグルはサンのJavaAPIから宣言コードをコピーした(同判決第106~107ページ)。今説明した通り、この事は、これらの37のパッケージについて、グーグルは必然的に個別のタスクに与えられた名前とパッケージとクラスにおけるこれらのタスクのグループ分けをコピーした事を意味する。

(付録Bのダイヤグラム表)

Javaapi_diagram

 さらに、上の概要の判示(a)の様に、ここで宣言コードは著作権保護対象と仮定してフェアユースについて判断するとして、宣言コードが著作権の保護対象か否かに関する明確な判示を避けている事にも留意が必要だが、その上で、アメリカ最高裁は、フェアユースの主要4要素についてそれぞれ判断し、特に、上の概要の(d)(2)の様に、第2要素の利用の目的及び性質について、プログラマーによる利用まで考慮して変形的利用であると、(d)(3)の様に、第3要素の利用部分の量及び本質性について、宣言コードは本質的なものとしてその変形的利用に結びついていると認め、(d)(4)に書かれている通り、第4要素の市場への影響について、APIの囲い込みが公衆の創造性を害する恐れも認め、この様な宣言コードのコピーはフェアユースであると判断しているのは、フェアユースについて一歩踏み込んだ判決と言っていいのではないかと思う。

 本判決のポイントは上記の概要だけでもほぼ掴めるのではないかと思うが、さらに補足として、第3要素と第4要素について書かれている本文から、それぞれキーとなる部分を抜き出して訳出しておくと、以下の様になるだろう。

We do not agree with the Federal Circuit's conclusion that Google could have achieved its Java-compatibility objective by copying only the 170 lines of code that are "necessary to write in the Java language." 886 F. 3d, at 1206. In our view, that conclusion views Google's legitimate objectives too narrowly. Google's basic objective was not simply to make the Java programming language usable on its Android systems. It was to permit programmers to make use of their knowledge and experience using the Sun Java API when they wrote new programs for smartphones with the Android platform. In principle, Google might have created its own, different system of declaring code. But the jury could have found that its doing so would not have achieved that basic objective. In a sense, the declaring code was the key that it needed to unlock the programmers' creative energies. And it needed those energies to create and to improve its own innovative Android systems.

We consequently believe that this "substantiality" factor weighs in favor of fair use.

私たちはグーグルがそのJava互換の目標を「Java言語で書くのに必要な」170コード行のみをコピーする事で達成し得たとする連邦裁判所の結論に同意しない(連邦裁判所判決第1206ページ)。私たちの考えでは、この結論は、グーグルの正当な目標を狭く見過ぎている。グーグルの基本的な目標は単にJavaプログラミング言語をそのアンドロイドシステムで利用可能とする事にあるのではない。それは、プログラマーがアンドロイドプラットフォームでスマートフォンのために新しいプログラムを書く時にサンのJavaAPIを用いていた時の知識と経験を利用可能とする事にある。原理的には、グーグルは宣言コードについて自ら異なるシステムを作り得たであろう。しかし、そうするのはその基本的な目標を達成するものではないと陪審員は判断しただろう。宣言コードはプログラマーの創作的エネルギーを解除するために必要なキーである。そして、自身の革新的なアンドロイドシステムを作り出し、改善するにはそのエネルギーが必要だったのである。

私たちは結果としてこの「本質性」の要素がフェアユースに有利に働くと考える。(第29~30ページ)

Finally, given programmers' investment in learning the Sun Java API, to allow enforcement of Oracle's copyright here would risk harm to the public. Given the costs and difficulties of producing alternative APIs with similar appeal to programmers, allowing enforcement here would make of the Sun Java API's declaring code a lock limiting the future creativity of new programs. Oracle alone would hold the key. The result could well prove highly profitable to Oracle (or other firms holding a copyright in computer interfaces). But those profits could well flow from creative improvements, new applications, and new uses developed by users who have learned to work with that interface. To that extent, the lock would interfere with, not further, copyright's basic creativity objectives. See Connectix Corp., 203 F. 3d, at 607; see also Sega Enterprises, 977 F. 2d, at 1523-1524 ("An attempt to monopolize the market by making it impossible for others to compete runs counter to the statutory purpose of promoting creative expression"); Lexmark Int'l, 387 F. 3d, at 544 (noting that where a subsequent user copied a computer program to foster functionality, it was not exploiting the programs "commercial value as a copyrighted work" (emphasis in original)). After all, "copyright supplies the economic incentive to [both] create and disseminate ideas," Harper & Row, 471 U. S., at 558, and the reimplementation of a user interface allows creative new computer code to more easily enter the market.

最後に、サンのJavaAPIの学習に対するプログラマーたちの投資があるとして、ここで、オラクルの著作権のエンフォースを認める事は公衆に害をなす恐れがあるだろう。プログラマーに対して似た訴求力を持つ代替APIを作成する事につき費用と困難性があるとして、ここで、エンフォースメントを認める事はサンのJavaAPIの宣言コードを新しいプログラムの未来の創作性を制限するロックとしてしまうだろう。オラクルだけがその鍵を持つ事になるだろう。この結果はオラクル(又はコンピュータインターフェースの著作権を有する他の企業)に高い利益をもたらす事だろう。しかし、この利益は、そのインターフェースで働く事を学んだユーザによって開発された創作的改善、新しいアプリケーション及び新しい利用から流入して来るものだろう。その限りで、このロックは、それ以上行かずとも、著作権の基本的な創作の目的を阻害する。コネクティックス事件判決第607ページ参照;セガ事件判決第152ページ参照(「競争する事を他者に不可能にする事によって市場を独占しようとする試みは創造的表現を促進するという法定の目的に反する」)。レックスマーク事件判決第544ページ参照(後のユーザが機能を育てるためにコンピュータプログラムをコピーした場合、それはプログラムの「著作物としての商業的価値」を利用したのではないと指摘)。結局の所、「著作権はアイデアの創作と普及[の両方]に経済的インセンティブを与える」(ハーパー&ロー事件判決第558ページ)ものであり、ユーザインターフェースの再実装は、より容易に市場に入る事を可能とする、新しい創作的コンピュータコードを可能とするのである。(第34~35ページ)

 このケースは、宣言コードは著作権保護対象でないとする事によっても同じ著作権非侵害の結論を出せたとは思うが(地裁判決ではその様な判断がされていた事もある)、そうしてしまうと、今度は自ら実行コードを書けば誰でもプログラミング言語の互換プラットフォームを提供できるという事になり、それはそれで逆の影響が大きく出かねないので、アメリカ最高裁は、著作権保護対象性についての判断を避け、グーグルのアンドロイドプラットフォームにおける部分としてのJavaAPIの宣言コードの利用についてフェアユースが適用されるとだけ言ってぎりぎりの解決を図ったものとも見られる。

 判決本文の細部や反対意見の訳出はここでは省略するが、この裁判が10年近くの長期に渡り裁判所間で行き来を繰り返している事や、一人の判事がその反対意見において同じ事実から真逆の著作権侵害の結論を導き出している事などからも、このケースはかなり微妙なものと知れる。その中で、この判決は、上で書いた通り、その部分だけを取り出して見れば、本質的な部分であるAPIの宣言コードの本来的な利用であるにも関わらず、ユーザ・利用者であるプログラマーによる利用まで考慮して変形的利用と捉え、APIの囲い込みが公衆の創造性を害する恐れも認めてフェアユースであるとした点で非常に興味深いものである。

 この判決が結論部分で述べている通り、確かに「コンピュータプログラムが主として機能的なものであるという事実は伝統的な著作権の概念をこの技術的な世界に適用する事を困難にしている」(上の概要部分の(e)又は本文第35ページ)のであって、そもそも著作権は、プログラミング言語のプラットフォームによる囲い込み、すなわち、プログラムやプログラマーそのものの技術的な囲い込みを正当化するべきものではなく、その点でアメリカ最高裁はフェアユースの意義から正しく解釈して結論を出していると私は思う。

 また、このケースはアメリカにおける大手ITプラットフォーマー間の限界紛争事例とも言え、その限りにおいて日本の法律問題に直結するとは思えないが、ひるがえって、日本法ではどうかと考えると、私自身は、この様なプログラム言語の宣言コードは著作物でないか著作権法第10条で著作権法の保護が及ばないとされているプログラム言語の一部であると考え、仮に保護を受ける場合でも第30条の4の思想感情非享受利用の権利制限の適用があると考えるが、それぞれ微妙な所があり、これらについて違う考えを持つ者もいるのではないかと思う。(第30条の4の権利制限の条文とその2018年の法改正経緯については第389回参照。)

 このケースのような限界事例のみをもって日本法についてどうこうと言うつもりは全くないが、日本においても、様々な判例からアメリカ型のフェアユースの意義に関する議論がさらに深められる事を、技術的な利用だけであれば今の第30条の4でそれなりの部分がカバーされているとも思えるが、そこに留まらず、より広く公正利用と考えられる変形利用やビジネス利用も含め著作物の公正利用を促すためにアメリカ型のフェアユースを導入する事が重要であるという考えが広く浸透する事を私は常に願っている。

(2021年5月25日夜の追記:幾つか誤記を直し、合わせ少し文章を整えた。)

| | コメント (0)

2016年2月21日 (日)

第358回:法定損害賠償制度の見直しを求めるアメリカ商務省の報告書

 先週金曜2日19日の衆議院予算委員会での民主党の福島伸享議員の質問によってにわかにTPP協定により導入を求められている法定損害賠償が話題となっているが(衆議院インターネット中継huffingtonpost.jpの記事ねとらぼの記事参照)、国会でのこのような議論を受け、場合によっては何かしらさらに国内でのTPP問題がややこしくねじれることもあるかも知れないと私は危惧している。

 アメリカでは、著作権法上の法定損害賠償として、その第504条第(c)項に著作物毎に通常750ドルから3万ドル、故意の場合はさらに15万ドルまで増額できるとして非常に大きな額が条文に書き込まれている(米著作権局HPの条文又は著作権情報センターHPの翻訳参照)。しかも、これは著作物一つあたりの額なので、多数の著作物が侵害される場合、個人には到底支払うことができない(恐らく企業であっても支払い難い)莫大な損害賠償額が請求され得るというのがその最大の問題である。

 その問題は私もパブコメ等で何度も指摘しているが、今まで前提をかなり飛ばして書いてしまっているところもあり、丁度良くつい最近の1月28日にアメリカ商務省の出したリミックス、ファーストセール及び法定損害賠償に関する白書(pdf)概要(pdf)リリース1も参照)はアメリカの制度とその問題を知る上で恰好の内容を含んでいるので、今回はこの報告書の内容を紹介する形で、その前提も含め、法定損害賠償の問題を取り上げたいと思う。

 なお、この報告書中のリミックスやファーストセールドクトリンの話も重要でないということはないのだが、それぞれ、リミックスについてはガイドラインを作るべきと、ファーストセールドクトリンについてはデジタル販売にまで拡張するのは時期尚早とそこまで踏み込んだ提言にはなっていないのでここでは省略し、一番大きな見直しを提言している法定損害賠償制度に関する部分のみ取り上げる。

 まず、この報告書の概要(pdf)から、法定損害賠償に関する部分を抜き出すと以下のようになる。(以下、例によって翻訳は全て拙訳。)

The Task Force is mindful that statutory damages serve a critical compensatory and deterrent function, and are particularly important in cases of online infringement, where the scope of the infringing use may not be ascertainable. At the same time, however, excessive and inconsistent awards can risk encouraging disrespect for copyright or chilling investment in innovation. The Task Force's inquiry focused on the appropriate calibration of statutory damages that may be assessed against individual file‐sharers and against online services, which can be secondarily liable for infringement of large numbers of works. It recommends the following three amendments to the Copyright Act to address some of the concerns presented:

  • Incorporate into the Copyright Act a list of factors for courts and juries to consider when determining the amount of a statutory damages award, providing a greater degree of predictability in infringement cases across the country. In considering what factors should be included, the Task Force drew upon existing model jury instructions as well as federal case law.
  • Amend the copyright notice provisions to remove a bar to eligibility for the lower "innocent infringement" statutory damages awards.
  • In cases involving non‐willful secondary liability of online services offering a large number of works, give courts discretion to assess statutory damages other than on a strict per‐work basis.

The Task Force also supports the creation of a streamlined procedure for adjudicating small claims of copyright infringement and believes that further consideration should be given to the proposal of the Copyright Office to establish such a tribunal. This could help diminish the risk of disproportionate levels of damages against individual file‐sharers.

タスクフォースは、法定損害賠償は重要な補償及び抑止機能を果たすものであり、侵害利用の範囲が不明確となり得るオンライン侵害の場合に特に重要であることに留意する。同時に、しかしながら、過大で一貫性のない賠償は著作権の軽視を助長したり、イノベーションへの投資を萎縮させたりする恐れがある。タスクフォースの調査は、個人のファイル共有と、多数の著作物の侵害について二次的な責任を負い得るオンラインサービスに課され得る法定損害賠償の適切な見直しに焦点をあてたものである。タスクフォースは、示された幾つかの懸念への対処として以下の3つの著作権法の改正を勧告する。

  • 法定損害賠償の額を決定する際に裁判所と陪審員が考慮するべき要素のリストの著作権法への導入、これは全国で侵害事件における予見可能性をより高めるものである。どのような要素が入れられるべきかを検討するにあたり、タスクフォースは既存のモデル陪審説示並びに連邦判例法を利用した。
  • 「罪のない侵害」の場合の法定損害賠償の減額に対する適格性の障害を除去するための著作権ノーティス規定の改正。
  • 多数の著作物を提供するオンラインサービスの故意でない時の二次責任に関する場合に、厳密に著作物数を基礎とする以外に法定損害賠償を評価する裁量を裁判所に与えること。

タスクフォースはまた著作権侵害の少額訴訟を裁定するための簡素な手続きの創設を支持し、さらなる検討がそのような裁定所を確立する著作権局の提案のために加えられるべきであると考える。このことは個人のファイル共有に対する不釣り合いな損害賠償の危険を解消することに役立つであろう。

 ここで、アメリカ商務省が「過大で一貫性のない賠償は著作権の軽視を助長したり、イノベーションへの投資を萎縮させたりする恐れがある」と認め、法定損害賠償を制限する方向への法改正を勧めていることの意味は大きいだろうし、この概要部分だけでも簡にして要を得ていると思うが、さらに詳しく見るために、報告書本文(pdf)から、ポイントとなるを抜き出して行くと、まず、個人によるファイル共有侵害との関係における問題について、70ページ以降に以下のように書かれている。

a. The Level of Statutory Damages

The Task Force heard differing viewpoints about whether the full range of statutory damages permitted under current law should be applied to individual file-sharers. Many commenters addressed the perceived fairness of enforcement activities vis-a-vis individual defendants generally, including whether the possibility of large statutory damages awards impedes individuals' ability to defend themselves or gives rise to abusive tactics. The Task Force also received proposals that addressed other concerns about litigation against individual file-sharers.

Much of the discussion focused on the only two such cases with reported trial verdicts, both of which resulted in high statutory damages awards that received considerable publicity.

Those favoring an adjustment of the permissible range of statutory damages pointed to the large jury awards in these cases as examples of the "unpredictability and irrationality" of the statutory damages regime. According to those commenters, individual file-sharers who infringe a handful of works for private, non-commercial purposes should not be required to pay damages that are disproportionate to the market value of the works. Nor, in their view, does it make sense to impose large damages awards on individuals who cannot pay anywhere near the amounts awarded. Some argued that the awards in these cases "exceed[ed] any rational measure of deterrence," noting potential due process concerns.

...

b. Inconsistencies in Application

A number of commenters asserted that because there is no specific set of factors or guidelines to be used in calculating statutory damages awards, troubling inconsistencies in their levels can arise. One commenter pointed out that damages are "untethered from anything." While few examples involving file sharing were offered, the four different verdicts in the two cases discussed above represent a wide array of awards based upon similar or identical facts.

Copyright owner groups as well as consumer advocates and other stakeholders suggested that it would be helpful to provide courts with guidance on factors to consider when setting statutory damages awards. Several noted that although some model jury instructions already exist, they are not being used in a consistent manner or in every circuit.

c. Litigation Abuse

Several commenters claimed that the potential high levels of statutory damages are linked to troubling enforcement tactics, often at the expense of individual defendants. In their view, the statutory damages regime has helped foster a "nationwide plague of lawsuit abuse over the past three years" by entities that they label "copyright trolls." They pointed to reports that attorneys representing copyright holders have filed hundreds of lawsuits against tens of thousands of anonymous Internet users, representing a substantial percentage of copyright suits filed in the federal courts in recent years. According to the commenters, these cases are rarely if ever litigated; instead, attorneys file "boilerplate complaints based on a modicum of evidence, calculated to maximize settlement profits by minimizing costs and effort" and use the courts' subpoena power to identify Internet users, often in multiparty John Doe actions. Then, commenters allege, they engage in a campaign of threats of high potential damages and harassment to coerce their targets into paying settlements of $2,000 to $10,000. This is a particular concern with respect to suits about copyrights covering adult content, where the TaskForce has noted a large number of settlements that may have been motivated in part by the defendants' desire to avoid having their names associated with such content. Such "predatory" enforcement may contribute to a negative public image of copyright.

This behavior was characterized as the product of a "litigation business model where a plaintiff uses copyright law ‘not to protect its property from unlicensed use, but rather to generate profit from use even in the absence of articulable harm to' the plaintiff." One commenter stated that"[h]olders of low-value copyrights in unsuccessful movies or low-cost pornography, and even invalid assignments of rights in newspaper articles, use the threat of statutory damages to turn litigation threats into a profit center." Another predicted that if statutory damages were calibrated to much lower levels, then this behavior would disappear.

...

a.法定損害賠償の水準

タスクフォースは、現行法において許されている法定損害賠償の全範囲が個人のファイル共有に適用されるべきかということについて様々な見解を聞いた。多くの意見は、巨額の法定損害賠償の可能性が個人の被告の自己弁護能力を阻害しているか、濫用戦略をもたらしているかについても含め、一般的に個人のファイル共有に対する権利行使について知覚される公平性に向けられていた。タスクフォースは、個人のファイル共有に対する訴訟に関する他の懸念に対処する提案も受けた。

議論は多く、ともに大きく喧伝されている、高額の法定損害賠償をもたらした、報告された判決のあったそのような2つのケースにほぼ集中していた。

彼らは法定損害賠償のあり得る範囲の適正化を支持し、法定損害賠償制度の「予見不能性と不合理性」の例としてこれらのケースにおける巨額な陪審員賠償をあげた。これらの評者によれば、私的に非商業目的で一握りの著作物を侵害する個人のファイル共有について、その著作物の市場価値と不釣り合いな損害賠償を支払うことは求められるべきではない。また、彼らの見解では、如何なる場合でもそのような損害賠償額をほぼ支払うことができない個人に高額の損害賠償を課すのは馬鹿げている。何人かは、これらのケースにおける賠償は「抑止の合理的な手段を超えている」と主張し、潜在的な適正な法的手続きに対する懸念を示した。

(略)

b.法適用における一貫性のなさ

数多くの評者が、法定損害賠償を算定するための特定の要素一式又はガイドラインがないことから、その水準における問題の多い一貫性のなさが生じ得ると主張している。ある評者は、損害賠償は「あらゆることから解き放たれている」と指摘している。ファイル共有がからむ例は多く提供されていないが、上記の2つのケースにおける4つの異なる判決が同様の又は同一の事実に基づく賠償の広い幅を示している。

著作権者のグループ並びに消費者の代表及び他の利害関係者は、法定損害賠償を決定する際に考慮する要素についてのガイドを裁判所に提供することが助けになると示唆している。何人かは、モデル陪審説示が既に存在しているものの、それは一貫性のある形であるいはあらゆる法廷で使われていないと注意している。

c.訴訟の濫用

何人かの評者は、潜在的に高い水準の法定損害賠償が問題の多い権利行使戦略に結びついており、個々の被告に被害を与えていることも多いと訴えている。彼らの見解では、法定損害賠償制度は、彼らが「著作権トロール」と名づける主体による「3年来の全国での訴訟の濫用の大発生」を助長して来た。彼らは、著作権者を代理する弁護士が何万人という匿名のインターネットユーザに対して何百という訴訟を提起し、これが近年連邦裁判所に提起された著作権訴訟のかなりの割合を占めている可能性があるとの報告があることを指摘している。これらの評者によれば、これらのケースにおいて訴訟が遂行されることはまれである。そうではあるが、弁護士たちは「費用と労力を最小化しながら最大の和解の利益を受けられると見込んで僅かな証拠に基づいて定型の訴えを提起しており」、インターネットユーザを特定する裁判所の召喚権限を用い、これは多数の関係者を含むジョン・ドウ(名なし)裁判によることも多い。すなわち、これらの評者の主張するところ、彼らはあり得る高額の損害賠償により脅迫し、嫌がらせによって相手を2千ドルから1万ドルの和解金を支払うよう追い込むキャンペーンに従事している。このことはアダルトコンテンツを含む著作権に関する訴訟に関して特に懸念される、タスクフォースは、そこではそのようなコンテンツと自身の名前が結び付けられることを避けたいと思う原告の望みがその動機の一部になっている多数の和解があると注意する。このような「人を食い物にする」権利行使は著作権の否定的なイメージの流布に寄与しているであろう。

この振る舞いは、「原告が『不正な利用からその所有権を保護するためではなく、どちらかと言えば、原告に明確に言える害が存在しない場合における利用から利益を生み出すため』に著作権法を利用している訴訟ビジネスモデル」が作り出したものとして特徴づけられる。ある評者は、「成功しなかった映画又は低費用のポルノグラフィーにおける価値の低い著作権の所有者が、また新聞記事の無効な権利譲渡によってすら、法定損害賠償による脅迫を利用して訴訟による脅迫をプロフィットセンターとしている」と述べている。他のある評者は、法定損害賠償が遥かに低い水準に補正されていれば、この振る舞いは消えるだろうと予測している。

(略)

 次に、二次的な責任、つまり、インターネットサービスプロバイダーの間接侵害責任との関係における問題ついて、79ページ以降に以下のように書かれている。

a. Chilling Effects

With respect to online service providers, the question posed by the Task Force was how statutory damages should be calculated in cases involving secondary liability where hundreds of thousands of works may have been infringed. In addition to concerns expressed about statutory damages generally, the comments most relevant to this question focused on whether potentially huge statutory damages awards have a "chilling effect" on innovation and investment.

Critics of the current statutory damages regime pointed to a number of lawsuits against technology companies, generally involving services offering methods of digital distribution that enable large-scale copyright infringement by third parties. In determining the responsibility of these companies for the illegal activities of the users, the courts have applied various theories of direct and secondary liability. Although many of these cases ultimately settled, the levels of potential liability have fueled headlines and commentary.

Technology companies and public interest advocates asserted that the magnitude of statutory damages awards available in such cases have had a chilling effect on innovation and investment. One commenter cited an amicus curiae brief to the Supreme Court by a group of venture capitalists arguing that statutory damages have "crushing implications for vendors of multi-purpose technologies, where damages from unforeseen users can quickly mount in the millions or even billions of dollars[,]" and that in this respect secondary liability for copyright infringement is "qualitatively different from most other sorts of business risks that investors can insure against or build into their business calculations." Another commenter pointed to a survey of investors in digital content intermediaries that it said confirmed "that uncertainty around liability risks deter[ring] investment in this field."

Commenters also pointed to innovative companies that they say were bankrupted by litigation even though they were ultimately found to be non-infringing. According to one commenter, potential statutory damages will deter some new business plans that rely on fair use from moving forward. At the same time, several noted that "[e]vidence for the innovation-chilling effect will ... usually not be readily apparent. In most cases, the public doesn't see and will likely never know about the innovations that don't happen and the features that aren't offered."

...

a.萎縮効果

オンラインサービスプロバイダーに関して、タスクフォースが出した質問は、何十万の著作物が侵害される二次責任が絡む場合にどのように法定損害賠償が算定されるべきかというものである。法定損害賠償について一般的に示された懸念に加えて、ほぼこの質問に関する回答は、莫大なものとなり得る法定損害賠償がイノベーションと投資に対し「萎縮効果」を有するかどうかということに集中していた。

現行の法定損害賠償制度の批判者は、一般的に第三者による大規模な著作権侵害を可能とするデジタル頒布手段を提供するサービスに関わるテクノロジー企業に対する数多くの訴訟があることを指摘している。これらの企業のそのユーザの違法な活動に対する責任を決定する際、裁判所は直接及び二次責任の様々な理論を適用して来た。これらのケースの多くが最終的に和解しているものの、あり得る責任の水準が見出しと論評に燃料を提供して来た。

テクノロジー企業と公共の利益の擁護者は、このような場合に適用され得る多額の法定損害賠償はイノベーションと投資に対し萎縮効果を有していると主張している。ある評者は、法定損害賠償は「予見できないユーザによる損害賠償が即座に何百万ドルか、何十億ドルかにも積み上がり得る、多目的技術の販売業者にとって破壊的な意味を持ち得」、この点で著作権侵害の二次責任は「投資家がそれに対して保険を掛け得るか、そのビジネス勘定の中に組み入れることのできる他のほとんどのビジネスリスクと質的に異なる」と主張するベンチャーキャピタリストのグループから最高裁判所へのアミカスブリーフを引用していた。他のある評者は、「責任リスクに関する不確実性がこの分野への投資を抑止している」のは確かであると言っているデジタルコンテンツ仲介者についての投資家の調査があることを指摘している。

評者は、イノベーション企業は最終的に非侵害であるとされたとしても訴訟で破産し得ると指摘している。ある評者によれば、あり得る法定損害賠償は、フェアユースに依拠する新たなビジネスプランを前に進めることを抑止することもあるであろう。同時に、何人かは、「イノベーションの萎縮効果についての証拠は…通常簡単に出て来ない。ほとんどの場合において、公衆は、起こらなかったイノベーションと提供されなかった機能を見ることはなく、それについて知ることも決してないであろう」と指摘している。

(略)

 最後に、これに対する解決策の提言としては、第85ページ以降に、以下のように書かれている。

1. Overview

In reviewing the positions and proposals set forth by stakeholders, the Task Force is mindful that statutory damages are intended to "provide reparation for injury" as well as to "discourage wrongful conduct." We agree that there is a need for effective enforcement tools, including meaningful statutory damages, to curb the online piracy that can undermine the value of rights and hobble the development of legitimate markets. At the same time, however, it is important to avoid excessive and inconsistent awards that risk encouraging disrespect for copyright law or chilling investment in innovation. And the abusive enforcement campaigns reported by commenters should not be tolerated.

Accordingly, the Task Force recommends three amendments to the Copyright Act to address some of the concerns presented and to better balance the needs of copyright owners, users, and intermediaries. First, we recommend incorporating into the statute a list of factors for courts and juries to consider when determining the amount of a statutory damages award. Second, we recommend changes to the copyright notice provisions that would expand eligibility for the lower "innocent infringement" statutory damages awards. We also propose that, in cases involving non-willful secondary liability for online services offering a large number of works, courts be given discretion to assess statutory damages other than on a strict per-work basis. Together, these changes should maintain the goals of compensation and deterrence that thestatutory damages regime supports while providing courts with improved tools to appropriately calibrate the awards.

2. Recommendations

a. Specify Factors in the Copyright Act to Consider in Assessing Statutory Damages

The Task Force recommends that Congress enact a new paragraph in Section 504 of the Copyright Act specifying factors that must be considered when determining statutory damage award amounts. The aim is to ensure a greater degree of predictability in copyright infringement cases across the country and address some other concerns raised in this proceeding. In considering what factors should be included, we have drawn upon existing model jury instructions as well as federal case law. The Task Force considered proposing federal model jury instructions, but concluded that a statutory set of factors would be preferable since they will be binding on all courts. We believe that litigants and courts would be well-served by requiring consideration of a uniform set of factors designed to result in an appropriate award based upon the facts of each case.

The nine factors listed below are those that will most often be applicable in a statutory damages determination. We believe that they should be non-exclusive, so that courts are not foreclosed from considering other factors that may be relevant in a particular case.

The Task Force proposes a new clause in subsection Section 504(c) as follows:

FACTORS TO CONSIDER -- In making any award under this subsection, a court shall consider the following nonexclusive factors in determining the appropriate amount of the award:

(1) The plaintiff's revenues lost and the difficulty of proving damages.

(2) The defendant's expenses saved, profits reaped, and other benefits from the infringement.

(3) The need to deter future infringements.

(4) The defendant's financial situation.

(5) The value or nature of the work infringed.

(6) The circumstances, duration, and scope of the infringement, including whether it was commercial in nature.

(7) In cases involving infringement of multiple works, whether the total sum of damages, taking into account the number of works infringed and number of awards made, is commensurate with the overall harm caused by the infringement.

(8) The defendant's state of mind, including whether the defendant was a willful or innocent infringer.

(9) In the case of willful infringement, whether it is appropriate to punish the defendant and if so, the amount of damages that would result in an appropriate punishment.

When calculating the total award, all of these factors should be weighed holistically, in the context of the entire case, to ensure that the overall award is appropriate. Below we explain various factors' relevance to the issues raised in the comments and roundtables.

...

1.概観

利害関係者が示した位置と提案を検討するにあたり、タスクフォースは、法定損害賠償は「侵害に対する補填を提供する」とともに「不正行為を阻止する」ことを目的としていることに留意する。我々は、権利の価値を下げ、適法な市場の発展を阻害するオンライン海賊行為を抑制するために意味のある法定損害賠償を含む、有効な権利行使ツールの必要性があることを認めている。同時に、しかしながら、著作権の軽視を助長したり、イノベーションへの投資を萎縮させたりする恐れがある過大で一貫性のない賠償を避けることは重要である。そして、評者の報告する濫用的な権利行使キャンペーンは許容されるべきでない。

したがって、タスクフォースは、示された幾つかの懸念に対処し、著作権者、利用者及び仲介者の必要性をより良くバランスさせるために3つの著作権法改正を勧告する。第一に、法定損害賠償の額を決定する際に裁判所と陪審員が考慮するべき要素のリストの著作権法への導入を勧告する。第二に、我々は、「罪のない侵害」の場合の法定損害賠償の減額に対する適格性の障害を除去するための著作権ノーティス規定の改正を勧告する。我々はまた、多数の著作物を提供するオンラインサービスの故意でない時の二次責任に関する場合に、厳密に著作物数を基礎とする以外に法定損害賠償を評価する裁量を裁判所に与えることも提案する。合わせ、これらの修正は、賠償を適正に補正する改善されたツールを裁判所に与えつつ補償と抑止の目標を維持するべきである。

2.勧告

a.法定損害賠償の評価において検討するべき要素の著作権法における特定

タスクフォースは、議会が、著作権法の第504条において法定損害賠償額を決定する際に検討されなくてはならない要素を特定する新たな項を立法することを勧告する。その目的は、全国での著作権侵害事件における予見可能性の向上を確保し、本手続きにおいて提起された他の幾つかの懸念に対処することにある。どのような要素が入れられるべきかを検討するにあたり、我々はは既存のモデル陪審説示並びに連邦の判例法を利用した。タスクフォースはモデル陪審説示の提案も検討したが、全ての裁判所を拘束するものとなることから要素一式の法定の方が好ましいと結論づけた。それぞれのケースの事実に基づく適切な賠償をもたらすよう適切に設計された統一的な要素一式の考慮を求めることは訴訟当事者と裁判所に非常に役立つことであろうと我々は考えている。

以下にあげる9つの要素は法定損害賠償の決定において最も良く適用されるだろうものである。個別のケースにおいて関係して来得る他の要素を考慮することを排除されないよう、これらは非排他的なものであるべきと我々は考えている。

タスクフォースは第504条第(c)項に以下のような新条項を入れることを提案する。

考慮要素 − 本項の賠償を課すにあたり、裁判所は賠償の適正な額を決定する際に以下の非排他的要素を検討しなければならない:

(1)原告の逸失収入及び損害を証明する困難性。

(2)被告の低減された費用、かすめ取った利益及び侵害から来る他の便益。

(3)将来の侵害の抑止の必要性。

(4)被告の経済的状況。

(5)侵害された著作物の価値又は性質。

(6)それが商業的な性質を持つかどうかも含め、侵害の状況、継続性及び範囲

(7)複数の著作物の侵害が絡む場合において、侵害された著作物の数及び課される賠償の数を考慮に入れ、損害賠償の総額が侵害によって引き起こされた害全体に釣り合っているか。
(8)原告が故意か罪のない侵害者であるかも含め、被告の心理状態。

(9)故意の侵害の場合に、被告を罰するのが適切であるかどうか、そうであるとして、適正な罰となる損害賠償額。

全賠償額を算定する際、賠償全体が適正なものとなることを確保するよう、これらの要素の全てが、総合的に比較衡量されるべきである。以下で我々は意見と討論で提起された問題と様々な要素との関係を説明する。

(略)

 なお、報告書の第71ページの注414に、個人が被告となり、大きく報道された訴訟における損害賠償額として、タンネンバウム事件の67万5千ドル(侵害されたのは30著作物で、1著作物あたり2万2千500)、3つのトーマス事件でそれぞれ22万2千ドル、192万ドル、150万ドル(1著作物あたりそれぞれ9千250ドル、8万ドル、6万2千五百ドル)といった巨額の数字もあげられており、第94ページの注559に、違法ファイル共有に対する法定損害賠償の額として1著作物あたり典型的には750ドルと6千500ドルの間になっているという数字もあげられている。またなお、同じ注414に、多くのケースにおいて訴訟が法定損害賠償額の決定まで行くことは少ないとも書かれ、和解金の範囲は大体3千ドルから5千ドル程度か2千ドルから1万ドル程度とも書かれているが、このような和解金ですらそれなりに大きな額だろう。

 このような報告書を読めば、アメリカの法定損害賠償制度が実際の損害としてはあり得ないほど巨額のデタラメな損害賠償請求を可能としており、著作権トロールによる訴訟の濫用を招き、社会的混乱の原因となっていることが分かるだろうし、インターネットサービスプロバイダーとの関係でイノベーションに対する萎縮効果が生じている可能性が指摘されていることも分かるだろう。アメリカにおいても、このような批判から、法定損害賠償制度について日本における填補賠償の考え方(損害賠償は実際の損害額を補填するものであるという考え方)に近づける形で見直すべきという提案が政府の省庁レベルで出されている状況にあるのである。(ただし、アメリカのことなので、日本と違って、政府省庁が報告書を作ったからと言ってそのまま法改正がされるようなことはほぼないが。)

 そして、TPP協定と法定損害賠償の関係についてだが、TPP協定の条文上、第18.74条の第6項から第8項により、「あらかじめ定められた賠償」すなわち法定損害賠償を選ぶ場合には、「将来の侵害を防ぐ目的も含め、侵害から発生した損害に対し権利者を十分に補償するであろう額として設定されなければならない」とされているのはその通りととは言え、この条文に解釈の幅があることも間違いない。(TPP協定の条文は第532回の私の部分訳か、TPP政府対策本部知財章全文仮訳(pdf)参照。)

 このTPP協定の規定上の「将来の侵害を防ぐ目的」も含む法定損害賠償が本当にアメリカ型の懲罰的かつ抑止的なレベルの法定損害賠償制度を必ず意味するとしたならば、日本の民法上の填補賠償の考え方に全く合わないのは確かである。(前回取り上げた2月10日の文化庁の法制・基本問題小委員会環太平洋パートナーシップ(TPP)協定に伴う制度整備の在り方等について(案)(pdf)でも引用されている萬世工業事件最高裁判決(pdf)がこのような懲罰的又は抑止的損害賠償の考え方を完全に否定している。)

 しかし、前回も書いた通り、今のところの文化庁の整理では、このTPP協定の法定損害賠償規定はそこまで強力な法定損害賠償制度の導入を求めているものではなく、上記の法制・基本問題小委員会の文化庁資料(pdf)においても、「(著作権法の)第114条第3項については、権利者が侵害行為により実際に生じた損害額や損害と侵害行為との因果関係の立証をせずに、侵害者に対して使用料相当額という一定の範囲の額の支払を求める制度であり、上記の『法定の損害賠償』の定義に該当するとして、我が国は同項によって『法定の損害賠償』を担保しているとする考え方も必ずしも排除されない」(第31ページ)、「TPP協定の求める趣旨をより適切に反映する観点から、第114条第3項等の現行規定に加えて、填補賠償原則等の枠内で、実際に生じる損害との関係について合理的に説明が可能な額を法定する規定を別途設けることが適当」(第32ページ)、「著作権等管理事業者の使用料規程により算出した額を同項の使用料相当額として請求することができる旨を法律上明記することが適当」とされている。このような条文の明確化が侵害の抑止に全く結びつかないということもなく、文化庁の今の整理は国際条約の解釈として完全に無理があるということもないだろう。

 2月19日の衆院予算委での、TPP協定で求められている法定損害賠償制度は日本の国内法の損害賠償の考え方(填補賠償の考え方)と合わないのではないかという、TPP協定の条文を普通に読むならば当然想定される民主党の福島伸享議員の質問に対する岩城法務大臣の回答は全く要領を得ず、一面正しいながら今の日本の国内法の整理では損害賠償は抑止効果を持たないとすら言ってしまっているのだが、どうして現職の法務大臣が政府の公開検討の内容すら十分に把握できていないのか正直かなり理解に苦しむ。

 最終的にアメリカから文句をつけられる可能性を排除できないとは言え、今ここでTPP関連法改正案をアメリカ型の強力な法定損害賠償制度に変更することは法案に極めて大きな問題を一つ追加することに他ならず、良いことは何もない。

 例えば、福島議員は質問の中で、通常の填補賠償の考え方に基づくと損害賠償が低くなる場合の例として同人誌が侵害された場合をあげているが、この法定損害賠償の議論と同人誌を守る守らないという議論はちょっと軸が違う話である。普通想定されるところの「同人誌」は一次創作(オリジナル)だけではなく二次創作(パロディ)も含むものだろうし、二次創作は親告罪であることがグレー化に役立っているものの著作権侵害側にも立っているので、本当に厳しい法定損害賠償を入れたらかえって二次創作にとって悪い結果を招くこともあり得るのである。(今更ながら、損害賠償が低くなる例としては、アメリカの報告書であげられているように、全く売れなかった低予算映画や小さな新聞記事などをあげた方が適切だったろう。)

 本当にアメリカ型の強力な法定損害賠償制度を導入したら、さらに違法ダウンロード・アップロード・ファイル共有やインターネットサービスにおける間接侵害との関係なども問題になり、日本でもアメリカ商務省の報告書で指摘されているような個人ユーザに対する巨額の損害賠償、著作権トロールによる訴訟の濫用、インターネットサービスプロバイダーの事業における萎縮効果などの問題が発生して来るに違いない。

 TPP協定そのものに私は反対だが、話を法定損害賠償に限るなら、今の文化庁の整理は悪くないと思っている。文化庁の法制・基本問題小委は次回2月24日に開かれる予定だが(文化庁のリリース参照)、国会での議論から変な方向にねじれて欲しくはない。法定損害賠償のようなかなりマニアックな問題がなかなか理解され難いのはやむを得ないこととは言え、前提と目的をはっきりさせた上で全体的なことを考えて政策判断はしてもらいたいと私は常に思っている。

 なお、単純化のために上では話を著作権法に限ったが、商標法についても同様に法定損害賠償の導入が議論されていることは要注意である(日経の記事参照)。特許庁の2月12日の産業構造審議会・知的財産分科会TPP協定を担保するための商標法改正について(pdf)によると、商標法における法定損害賠償について「商標の不正利用による損害の賠償を請求する場合において、当該登録商標の取得及び維持に通常要する費用に相当する額を損害額として請求できる規定を追加する」とされているところだが、TPP協定と特許法や商標法の改正検討案の関係についてはまた回を分けたいと思っている。

(2016年2月22日夜の追記:2カ所間違っていたリンクを修正し、合わせ少し文章を整えた。)

| | コメント (0) | トラックバック (0)