扎克伯格让Meta（原Facebook）下载近3000 部色情影片训练自家AI,让成人电影公司给告了！
Lawsuit Claims Meta Pirated Vast Numbers of Dirty Movies to "Train Its AI"
译文简介
俩大版权流氓PK，结果如何拭目以待。另：各社交平台针对该文相关评论即使没限流，也已关闭。仅有其中油管一疑似主要为律师间探讨的评论贴未限流。
正文翻译
Lawsuit Claims Meta Pirated Vast Numbers of Dirty Movies to "Train Its AI"
JUL 30, 10:16 AM EDT by NOOR AL-SIBAI
诉讼指控Meta盗版海量成人电影用于"训练其AI"
Mark Zuckerberg's Meta has been accused of pirating — we are not kidding — a whole adult bookstore's worth of dirty movies to train its artificial intelligence models.
First flagged by the blog TorrentFreak, a copyright company called Strike 3 Holdings and an adult film studio called Counterlife Media have filed suit against Meta, alleging that the tech giant torrented nearly 2,400 copyrighted skin flicks. All in the name of AI research, of course.
马克·扎克伯格（Mark Zuckerberg）的Meta公司被指控盗版,我们没有开玩笑,该公司下载了整整一个成人书店容量的色情电影，用来训练其人工智能模型。
此事最初由博客 TorrentFreak 披露，一家名为Strike 3 Holdings的版权公司和一家名为Counterlife Media的成人电影制片公司已对Meta提起诉讼，指控这家科技巨头通过“Bittorrent”的方式下载了至少2400部受版权保护的色情电影。当然，这一切都是以人工智能研究的名义进行的。
According to the suit, which was filed in California federal court, Strike 3 and Counterlife discovered by tracing dozens of IP and email addresses that Meta began downloading and seeding their content via Bittorrent way back in 2018. Because of the seeding, the lawsuit alleges, the company formerly known as Facebook also allegedly engaged in "methodical and persistent distribution of those works" to other parties — including, potentially, to minors.
"Defendant has continuously infringed Plaintiffs’ Works for years," the suit claims, "often infringing the very same day the motion pictures are released."
根据在加州联邦法院提起的诉讼，Strike3和Counterlife通过追踪数十个IP和电子邮件地址发现，Meta早在2018年就开始通过“BT”和做种（seeding）下载分享他们的内容。诉讼称，由于这种分享行为，这家前身为 Facebook 的公司还涉嫌 “有计划、持续地向其他第三方分发这些作品”，其中可能包括未成年人。
诉讼声称：“被告多年来持续侵犯原告影片，经常在影片发布当天就进行了侵权。”
Citing Richard Kadrey et al v. Meta Platforms, an ongoing lawsuit brought in 2023 by authors whose work Meta has since admitted to pirating, the adult companies said they were alxed to the Zuckerberg-owned tech monolith's torrenting activities in January 2025 through coverage of that lawsuit.
Using various infringement analysis and IP tracking tools, Strike 3 and Counterlife found that some 47 IP addresses associated with Meta — including "at least one residential IP address of a Meta employee" — had downloaded their copyrighted content. Stranger still, the way the data moved suggested "non-human patterns," and "the acquisition of this content [may have been] for AI training data."
这两家成人电影公司援引了2023年“理查德・卡德雷等人（注：均为畅销书作家）诉 Meta 平台公司” 一案 —— 在该案中，Meta承认盗版了一些作家的作品。他们表示，通过2025年1月对该案的相关报道，他们才注意到这家由扎克伯格掌控的科技巨头的非法下载活动。
通过各种侵权分析和 IP 追踪工具，Strike 3 和 Counterlife 发现约47个与 Meta 相关的 IP 地址（包括 “至少一个 Meta 员工的住宅 IP 地址”）下载了他们受版权保护的内容。更奇怪的是，下载方式以“非人类模式”方式呈现出，且 “获取这些内容 [可能是] 为了人工智能训练数据”。
Though there's no exact stated reason why Meta, which long ago banned nudity on its platforms, would want to pirate all that smut, Strike 3 and Counterlife wagered a guess in the suit.
"Plaintiffs’ Works provide natural, human-centric imagery, which shows parts of the body not found in regular videos, and a unique form of human interactions and facial expression," the lawsuit reads, employing some choice legalistic innuendo to talk around the subject at hand. "Plaintiffs' motion pictures contain extended scenes without director cuts which enable AI models to experience continuity in a way that cannot be derived from most television shows or mainstream motion pictures."
尽管长期以来Meta禁止在自家平台发布（人体）裸露内容，但为何要盗取这么多色情内容无法说明，不过Strike 3和Counterlife公司在诉讼中做出了推测。
诉讼中写道：“原告的作品提供了自然、以人为核心的影像，展示了常规视频中看不到的身体部位，以及独特的人类互动和面部表情形式。” 这段话用了一些精心选择的法律性暗示来回避谈论具体内容，“原告的影片包含未经导演剪辑的长镜头，这使得AI模型能以一种从大多数电视节目或主流电影中无法获得的方式理解连续性。”
In short, Strike 3 and Counterlife seem to be claiming that Meta might have used their content to train its AI video generators, like Meta Movie Gen, to recreate human movement in ways that other stolen data simply can't quite nail, if you catch our drift.
"Such models will eventually create identical content for little to no cost," the suit alleges. "This will effectively eliminate Plaintiffs' future ability to compete in the marketplace."
简而言之，Strike 3和Counterlife公司似乎在表明，Meta可能使用了他们的内容来训练其AI视频生成器（如 Meta Movie Gen），以便更精准地重现人类动作，而这是其他被盗数据无法做到的——如果你明白我们的意思的话。
诉讼称：“这类模型最终将能以极低甚至零成本生成相同的内容。这将有效地消除原告未来在市场上的竞争能力。”
Along with the deletion of any copyrighted and pirated content and an injunction to permanently bar Meta from torrenting its work again, Counterlife and Strike 3 are seeking damages of up to $150,000 per stolen video. With 2,396 pieces of content on the line, those damages could go as high as $359 million, TorrentFreak notes.
While it seems strange to consider a world where Meta was using all that raunchy content to train AI, the fact remains that the company has admitted to pirating other content. With this compelling evidence and Zuckerberg's unquenchable thirst to stay abreast of the latest tech trends, it seems conceivable that Meta actually did what's being claimed — even if it didn't necessarily plan to release the money shot, so to speak.
除了要求删除所有版权和盗版内容以及永久禁止Meta再次通过“BT”方式下载其作品的禁令外，Counterlife和Strike 3还要求对每部被盗视频最高15万美元的赔偿金。TorrentFreak指出，鉴于涉及2396部内容，赔偿金可能高达3.59亿美元。
虽然想到 Meta 会用这些低俗内容来训练人工智能似乎很离奇，但事实是，该公司已承认盗版过其他内容。有了这些令人信服的证据，再加上扎克伯格对紧跟最新科技趋势的无尽渴望，Meta 确实有可能做出了上述指控中的行为 —— 可以这么说，即便它未必真想发布那些 “‘露骨‘镜头”。
JUL 30, 10:16 AM EDT by NOOR AL-SIBAI
诉讼指控Meta盗版海量成人电影用于"训练其AI"
Mark Zuckerberg's Meta has been accused of pirating — we are not kidding — a whole adult bookstore's worth of dirty movies to train its artificial intelligence models.
First flagged by the blog TorrentFreak, a copyright company called Strike 3 Holdings and an adult film studio called Counterlife Media have filed suit against Meta, alleging that the tech giant torrented nearly 2,400 copyrighted skin flicks. All in the name of AI research, of course.
马克·扎克伯格（Mark Zuckerberg）的Meta公司被指控盗版,我们没有开玩笑,该公司下载了整整一个成人书店容量的色情电影，用来训练其人工智能模型。
此事最初由博客 TorrentFreak 披露，一家名为Strike 3 Holdings的版权公司和一家名为Counterlife Media的成人电影制片公司已对Meta提起诉讼，指控这家科技巨头通过“Bittorrent”的方式下载了至少2400部受版权保护的色情电影。当然，这一切都是以人工智能研究的名义进行的。
According to the suit, which was filed in California federal court, Strike 3 and Counterlife discovered by tracing dozens of IP and email addresses that Meta began downloading and seeding their content via Bittorrent way back in 2018. Because of the seeding, the lawsuit alleges, the company formerly known as Facebook also allegedly engaged in "methodical and persistent distribution of those works" to other parties — including, potentially, to minors.
"Defendant has continuously infringed Plaintiffs’ Works for years," the suit claims, "often infringing the very same day the motion pictures are released."
根据在加州联邦法院提起的诉讼，Strike3和Counterlife通过追踪数十个IP和电子邮件地址发现，Meta早在2018年就开始通过“BT”和做种（seeding）下载分享他们的内容。诉讼称，由于这种分享行为，这家前身为 Facebook 的公司还涉嫌 “有计划、持续地向其他第三方分发这些作品”，其中可能包括未成年人。
诉讼声称：“被告多年来持续侵犯原告影片，经常在影片发布当天就进行了侵权。”
Citing Richard Kadrey et al v. Meta Platforms, an ongoing lawsuit brought in 2023 by authors whose work Meta has since admitted to pirating, the adult companies said they were alxed to the Zuckerberg-owned tech monolith's torrenting activities in January 2025 through coverage of that lawsuit.
Using various infringement analysis and IP tracking tools, Strike 3 and Counterlife found that some 47 IP addresses associated with Meta — including "at least one residential IP address of a Meta employee" — had downloaded their copyrighted content. Stranger still, the way the data moved suggested "non-human patterns," and "the acquisition of this content [may have been] for AI training data."
这两家成人电影公司援引了2023年“理查德・卡德雷等人（注：均为畅销书作家）诉 Meta 平台公司” 一案 —— 在该案中，Meta承认盗版了一些作家的作品。他们表示，通过2025年1月对该案的相关报道，他们才注意到这家由扎克伯格掌控的科技巨头的非法下载活动。
通过各种侵权分析和 IP 追踪工具，Strike 3 和 Counterlife 发现约47个与 Meta 相关的 IP 地址（包括 “至少一个 Meta 员工的住宅 IP 地址”）下载了他们受版权保护的内容。更奇怪的是，下载方式以“非人类模式”方式呈现出，且 “获取这些内容 [可能是] 为了人工智能训练数据”。
Though there's no exact stated reason why Meta, which long ago banned nudity on its platforms, would want to pirate all that smut, Strike 3 and Counterlife wagered a guess in the suit.
"Plaintiffs’ Works provide natural, human-centric imagery, which shows parts of the body not found in regular videos, and a unique form of human interactions and facial expression," the lawsuit reads, employing some choice legalistic innuendo to talk around the subject at hand. "Plaintiffs' motion pictures contain extended scenes without director cuts which enable AI models to experience continuity in a way that cannot be derived from most television shows or mainstream motion pictures."
尽管长期以来Meta禁止在自家平台发布（人体）裸露内容，但为何要盗取这么多色情内容无法说明，不过Strike 3和Counterlife公司在诉讼中做出了推测。
诉讼中写道：“原告的作品提供了自然、以人为核心的影像，展示了常规视频中看不到的身体部位，以及独特的人类互动和面部表情形式。” 这段话用了一些精心选择的法律性暗示来回避谈论具体内容，“原告的影片包含未经导演剪辑的长镜头，这使得AI模型能以一种从大多数电视节目或主流电影中无法获得的方式理解连续性。”
In short, Strike 3 and Counterlife seem to be claiming that Meta might have used their content to train its AI video generators, like Meta Movie Gen, to recreate human movement in ways that other stolen data simply can't quite nail, if you catch our drift.
"Such models will eventually create identical content for little to no cost," the suit alleges. "This will effectively eliminate Plaintiffs' future ability to compete in the marketplace."
简而言之，Strike 3和Counterlife公司似乎在表明，Meta可能使用了他们的内容来训练其AI视频生成器（如 Meta Movie Gen），以便更精准地重现人类动作，而这是其他被盗数据无法做到的——如果你明白我们的意思的话。
诉讼称：“这类模型最终将能以极低甚至零成本生成相同的内容。这将有效地消除原告未来在市场上的竞争能力。”
Along with the deletion of any copyrighted and pirated content and an injunction to permanently bar Meta from torrenting its work again, Counterlife and Strike 3 are seeking damages of up to $150,000 per stolen video. With 2,396 pieces of content on the line, those damages could go as high as $359 million, TorrentFreak notes.
While it seems strange to consider a world where Meta was using all that raunchy content to train AI, the fact remains that the company has admitted to pirating other content. With this compelling evidence and Zuckerberg's unquenchable thirst to stay abreast of the latest tech trends, it seems conceivable that Meta actually did what's being claimed — even if it didn't necessarily plan to release the money shot, so to speak.
除了要求删除所有版权和盗版内容以及永久禁止Meta再次通过“BT”方式下载其作品的禁令外，Counterlife和Strike 3还要求对每部被盗视频最高15万美元的赔偿金。TorrentFreak指出，鉴于涉及2396部内容，赔偿金可能高达3.59亿美元。
虽然想到 Meta 会用这些低俗内容来训练人工智能似乎很离奇，但事实是，该公司已承认盗版过其他内容。有了这些令人信服的证据，再加上扎克伯格对紧跟最新科技趋势的无尽渴望，Meta 确实有可能做出了上述指控中的行为 —— 可以这么说，即便它未必真想发布那些 “‘露骨‘镜头”。
评论翻译
很赞 ( 3 )
收藏
Most litigious copyright troll vs most flagrant copyright infringer
最爱打官司的版权流氓 vs 最明目张胆的版权侵犯者
@davidharshman7645
I dunno...Google feels more flagrant.
不好说...感觉谷歌更嚣张。
@Gelondil
@davidharshman7645 Google may have greater volume of copyrightable data in their AI models, but I'm not sure their method of obtaining it was more flagrant than Meta's. For the most part Google already had all that data to begin with, well before the AI boom began, and didn't need to go torrenting it.
谷歌的 AI 模型里可能有更多受版权保护的数据，但要说获取方式比Meta更嚣张，我不确定。谷歌在 AI 热潮来临前就已收集了大量数据，根本没必要去用BT下载的方式扩充数据量。
@BAD_CONSUMER
Batman vs Joker
就像蝙蝠侠 vs 小丑
@darkarima
@Gelondil Exactly. For all that Strike 3 is repulsive for a large number of separate reasons… they have an argument that by torrenting it, Meta were also uploading their content to other people. Whether their use is "transformative" or not, they were distributing it to others.
完全同意。尽管 Strike 3 也不是什么好人…… 但他们说的倒是有点道理，Meta 用种子下载的同时，也把其中内容上传给了他人。无论其用途是否属于 “合理使用”，但确实在向他人分发了这些内容。
@commbs9838
"Heartwarming: The worst people you know are all fighting."
暖心：你所认识的俩大烂人掐起来了。
@lexslate2476
Heartbreaking: Worst copyright troll you know might have a point.
糟心：你认识的最糟的版权流氓，说不定这次他真占了点理。
@I_Am_Your_Problem
Except your entire society is propped up with them. So where do you stand....
问题是你们整个社会都靠他们支撑着。所以你站哪边~？
@Yuzral
Either Meta gets forced to stop training AI off people's videos without compensation (and presumably lobotomise said AIs to remove the offending training material...if that's even possible) or Strike 3 lose to a megacorp that has the resources to countersue and obliterate them. Either way the rest of us seem set to come out ahead in some fashion, so...pass the popcorn?
要么Meta被迫停止在无补偿的情况下用人们的视频训练AI（并还得给这些 AI “做脑叶切除”，删掉有问题的训练素材…… 前提是这真能做到），要么 Strike 3 输给有资源反诉并搞垮他们的巨头企业。可无论何种结果，我们普通人似乎多少都能沾点好处，所以…准备吃瓜吧~
@goldenalt3166
More likely strike 3 is paid off.
更有可能是Strike 3被收买封口了。
@0LoneTech
That's not how this sort of thing ends. The typical result is the megacorp buys the claimant, and apply the "we have a bigger warchest" strategy to their aggressions.
这种事通常不会那样结束。典型的结果是巨型企业收购了原告，并对他们的侵略行为采取“用钱砸”的策略。
@mattstorm360
@goldenalt3166 Hopefully that would open the door to others to claim copyright works used to train AI.
希望这能为其他人就用于训练AI的版权作品索赔打开大门。
@cha0sniper
Por que Los nos dos? My hope is on Meta losing, and Strike 3 going bankrupt shortly afterwards
为什么不两败俱伤呢？我巴不得Meta败诉，Strike 3 也赶紧破产。
@Queldonus
The best case scenario for all of us is that both these companies spend a large amount of time of money in court attacking each other, thus reducing the resources they have to do damage elsewhere in the world.
对我们所有人来说最好的情况是，这两家公司花大量时间和金钱在法庭上互撕，从而减少它们在世界其他地方搞破坏的资源。
@Danthemanwithaplan7
I never dreamed the day would come when I saw a Three Strikes filing and thought "No wait they have a point"
我从没想过有一天看会到Strike3告别人，居然会想 “等等，他们说得还算有理”。
@I_Am_Your_Problem
They don't have a point... Not a single legal pillar to stand on.
他们没理的…… 连一个站得住脚的法律依据都没有。
@eaglestdogg
@I_Am_Your_Problem illegally downloading a ton of videos to train your AI is 100% illegal and on top of that immoral
非法下载大量视频来训练 AI，这百分百违法，而且极不道德。
@SavageGreywolf
Which means that Meta and Strike 3 will both be losing money on it. It's a beautiful symphony
这意味着 Meta 和 Strike 3 都会在这事儿上损失金钱。多美妙的一出好戏啊。
@davidharshman7645
Gotta admit...that brief sounds like one of the most robust accumulations of evidence that an AI model company intentionally uses pirated material to train their models. I am impressed and have no problem with them going after Meta for that particular 'crime'
不得不承认…… 这份诉讼书里的证据，恐怕是目前为止最充分的证据，证明一家 AI 模型公司故意使用盗版材料来训练模型。我挺佩服的，他们针对 Meta 的这种 “罪行” 提起诉讼，我完全没意见。
@silverXnoise
So much societal damage in a single courtroom…incredible.
一个法庭里，竟然聚集了这么多社会危害…… 简直难以置信。
@phillipsusi1791
Isn't this the company that was sanctioned for posting their "content" and encouraging people to download it in order to sue them, thus coming to the court with unclean hands?
这家公司不就是那种，自己发布 “内容” 还鼓励别人下载，然后反手起诉人家，结果因为 “手欠”被处罚的公司吗？
@JonPMeyer
I imagine that neither party to this suit is going to be anxious to have the details of their legal positions and factual discovery be part of the publicly available docket, much as we might want it to be so. I would guess that we'll fairly quickly see a a shroud of secrecy descend over this docket. This will be fascinating, but frustrating, to follow. Thanks for the non-partisan approach to reporting this.
我猜这场诉讼的双方都不会急于让自己的法律立场和事实证据细节公之于众，尽管我们很想知道。估计这案子的卷宗很快会被“束之高阁”。跟进这事会很有意思，但也会令人沮丧，感谢你用不偏不倚的态度报道这事。
@vgamesx1
I would hope both lose in some way, no idea how, but that would be the best outcome.
我希望双方都以某种方式输掉，虽然不知道怎么输，但最好是两败俱伤。
@DarthAbyss316
Holy crap Leonard, how many lawsuits does that work out to be by Strike 3 Holdings a year??? Honestly, Strike 3 Holdings needs to just stop. It’s pathetic at this point. ♂️
好家伙，Strike 3 Holdings 一年到底要打多少场官司啊？？说实话，他们真该收手了。这时候还这样，也太不要脸了。
@Gelondil
19000 lawsuits for 8-10 years would be about 2000 per year.
8-10年1.9万起官司，那大概是每年2000起。
@danpowell806
Plot twist: so many torrents were downloaded independently by different Meta employees independently of their actual job duties, just for personal use.
剧情反转：这么多种子文件，其实是 Meta 的不同员工各自私下下载的，跟他们的实际工作毫无关系，纯粹是个人用途。
@MAlanThomasII
I agree that the partial summary judgement in Bartz v. Anthropic is instructive here: Downloading the works for AI training might be transformative fair use in general, but concomitant infringement based on how they did so—in that case, the maintenance of a pirated library for non-training purposes; in this case, the resharing via BitTorrent—is not. Setting aside my doubts about the identification of secret Meta download servers, I think that I wouldn't be surprised to see Strike 3 win on the distribution angle.
Or Strike 3 might at least accept a more-than-nuisance-value settlement, given how much they're assuming about what's going on and how much discovery it would take to prove it. It would be long and expensive for both sides to litigate, and copyright trolls are not known for going the distance, as you pointed out. Meta might want to pay above nuisance value to avoid a protracted investigation into their methods and avoid even an outside chance of a large judgement against them.
我认为“”巴茨诉 Anthropic 案“中的部分简易判决很有参考意义：一般来说，下载作品用于 AI 训练可能构成 “转化性合理使用”，但伴随有侵权行为 —— 比如该案中为非训练目的维护盗版库，或本案中通过 BitTorrent 二次分享 —— 则不适用。暂且不论我对 “Meta 秘密下载服务器” 这一认定的怀疑，我觉得 Strike 3 很可能在 “分发” 这一角度胜诉。
或者，考虑到 Strike 3 对提的那么多推测证据，且需要大量取证才能证实指控，他们至少可能拿到一笔远超 “骚扰赔偿” 的和解金。诉讼对双方而言都漫长且昂贵，而且正如你所说，版权流氓从不擅长打持久战。Meta 或许愿意支付高于 “骚扰赔偿” 的金额，以避免对其方法的长期调查，同时规避哪怕一丝可能面临巨额判决的风险。
@asdfasdfasdf212
Now Meta is going to go and assimilate our favorite copyright attorney into the Meta collective to learn how to beat Strike 3 holdings. We will miss you!
现在 Meta 怕是要把我们最擅长打版权案件律师吸收进他们的Meta 团体，学习怎么打败 Strike 3 Holdings。我们会想念你的！
@jamesowens7148
Meta dexed my account randomly after 10 years. Screw them.
我用的10年META账号莫名其妙地给删了，去TM的
@YdenMk-II
Was Strike 3 also the people who also uploaded the torrents of their works to sue people?
会不会是Was Strike 3自己上传的种子，目的就是去告下载者的公司？
@0LoneTech
Yes. Their business model relies on claiming their rights are infringed when they're actively distributing without any indication they'd reserved those rights. It's entrapment at best.
是的。他们的商业模式就靠这一套：自己主动分发内容，却不做任何 “保留权利” 的说明，然后反咬一口说被侵权。好听的说法是，这叫”钓鱼执法”。
@ribzer57
No, that was actually a law firm (Prenda Law) that ended up producing their own "materials" so they could drop their clients and get all the money.
不，那其实是普伦达律师事务所（Prenda Law，注：著名流氓律师事务所）干的事，他们最后甚至自己制作 “素材”，以便甩掉客户、独吞所有赔偿款。
@EnraEnerato
Germany has the infamous "Waldorf-Frommer" lawoffice, tehy are a menace!
德国有个臭名昭著的 “Waldorf-Frommer” 律师事务所，简直就是个祸害！
@charlesmanning3454
I wonder what would if instead of a 1 time payout, copyright holders were entitled to a license fee every time a modeled trained on their work was used?
Even if the fee was tiny, AI requires such vast quantities of data that there legal use would be impractical. Or maybe there would be public domain only AI, not as smart as piracy AI but legal.
我不禁设想，若版权持有者获得的不是一次性补偿，而是基于其作品训练的模型，每次使用按次收授权费会怎样？即便单笔费用微不足道，但人工智能所需的海量训练数据，恐怕会使合法授权变得不切实际。或许最终只会存在基于公共领域作品的AI系统，虽然不如盗版AI智能，但至少合法合规。
@Lyoishi
The primary reason you would train ai on adult content would be the mythical "adult content filter" that magically makes the internet safe etc. (This has many fundamental issues) but they may get other benefits depending on what their goals are.
用成人内容训练 AI 的主要理由，可能是为了打造传说中能 “净化网络” 的 “成人内容过滤”（这本身就有很多根本性问题），不过根据开发者的不同目标，他们或许还能从中获取其他收益。
@Locutus
Business models like Strike 3 should not be rewarded. Copyright protection should primarily benefit the original creator, and once it's sold or transferred, the copyright protection becomes weakened, and once sold to an entity that specialises in copyright collection like S3, copyright protection should be further weakened.
Companies like S3 have a role in society to exist, just not as a copyright troll.
像Strike 3这样的商业模式不应得到法律纵容。版权保护的核心价值本应是保障原创者的权益，因此当版权被转售或转让后，其保护力度就应当相应削弱；若版权最终流入S3这类专业版权追偿实体手中，法律保护更应进一步弱化。
这类公司的存在自有其社会价值，但绝不应以"版权流氓"的形式存在。
@lockedonlaw
The fair use argument is if a human can consume content and interpret it, an AI program can as well. But a human cannot consume a book or film, rewrite the exact same idea, and rebrand it as their own work, and neither can AI. As many are aware, there is a four prong test for fair use and the amount of a copyrighted work used is one of those prongs. At least in theory, 100% of the work is being used here for some purpose. I can see valid legal arguments on both sides.
“合理使用”的论点在于，如果人类可以消费内容并解读它，AI也可以。但人类不能看完一本书或一部电影后，以完全相同的想法重现出来，并将其重新包装为自己的作品，AI也不能。众所周知，合理使用有一个四要素测试（four prong test），其中一个要素就是受版权保护作品的使用量。至少从理论上讲，本案中作品的 100% 都被用于某种目的。我能看到双方都有合理的法律论据。
@LazyBunnyKiera
The law needs to be changed so any companies that Seed the torrents can't sue, since they're essentially giving it away.
法律需要修改，让任何为做种行为的公司都不能起诉，因为他们本质上就是在免费分享。
@JoeSmith-cy9wj
General information, and I know nothing about this new lawsuit, but I've seen musicians who claim they've been sued over their own material. Apparently someone can take a copyrighted work, feed it to a music AI, have it generate something similar, copyright that, and the sue the author over the original. It's apparently happened many times and some make it a business.
说个普遍情况，我对这个案件所知甚少，但我见过有音乐人声称自己因为自己的作品被起诉。显然，有人会拿受版权保护的作品训练音乐AI，让它生成类似的东西，然后给生成的内容申请版权，再去起诉原作品的作者。这种事显然发生过很多次，有些人还把这当成了一门生意。
@pbear6251
This sounds like a number of years ago when the big record companies went after Napster and pretty much put it out of business.
感觉叫像是几年前各大唱片公司起诉 Napster（注：Napster，基于P2P技术的MP3文件共享软件），差不多把它搞垮了的那件事。
@cr10001
These copyright trolls are misusing the legal system to threaten and blackmail individuals. I think they deserve the Luigi Mangione treatment.
这些版权流氓就是在滥用法律体系来威胁和敲诈个人。我觉得他们该受到路易吉·曼焦内（行刑）式的惩罚。
（注：Luigi Mangione ，路易吉·曼焦内，年初枪杀美国联合健康保险公司CEO布莱恩·汤普森的那位）。
@0LoneTech
Meta have published some of their paperwork including the exchange "hey, maybe we shouldn't be grabbing clearly illegal torrents for this work" and "shut up, do it anyway, it's fine" (paraphrased from memory).
Meta 公布过一些内部文件，其中有这样的对话：“嘿，也许我们不该为了这项工作去爬取明显非法的种子文件”。
“闭嘴，照做就是了，没事的”（记得是这个意思）。
@dlbiggins
"When you're using the legal system to make a profit"... "as a business model"...
Isn't that almost the definition of much of the US legal system?
“当你利用法律体系来盈利”…“作为一种商业模式”…
这难道不是美国法律体系的一大特点吗？
@The_Nametag
Fair use only applies to how the copyrighted material is USED, not how it is OBTAINED. You still have to be in legal possession of the material in the first place. You can't just steal a movie and then claim you're using it to make a satire, you have to legally have access to the movie in the first place. And that's what this law suit is about. Strike 3 Holdings may be a horrific abuser of copyright law... but this is one case they absolutely should win, and win big, if they are right that Meta was illegally downloading massive amounts of their material, regardless of what Meta was planning to DO with it.
合理使用（Fair use）只适用于受版权保护的影视作品是如何被“使用”的，而不是如何“获取”。你首先必须合法拥有该作品。你不能偷了一部电影，然后你说它是用它来制作讽刺作品的，你必须首先合法观看到那部电影。而这正是这场诉讼的关键。Strike 3 Holdings 可能是版权法滥用流氓…但如他们对Meta大量非法下载的指控属实，那他们肯定能赢，并且大获全胜，无论Meta想用它来“做”什么。
@SeanBZA
Wonder if Meta could say that S3 was deliberately placing material out there, thus actively relinquishing the copyright to that material. then they could be held liable for all cases they had filed for filing fraudulently, and that they had relinquished the copyrights by deliberately publicly sharing them.
想知道 Meta 会不会说，是Strike 3故意把这些影片放出去的，如果是，就相当于主动放弃版权。那样的话，Strike 3 就可能因为欺诈性诉讼而被追责，而且他们因为故意公开分享，就已经放弃了版权。
@Overonator
Let them fight. Both horrible companies.
让他们打去吧。这两家都不是什么好东西。
@Truth_Sapphire_Truth_Light_Tru
Copyright trolls vs AI scrapers
All I have to say is... Let them fight.
版权流氓VS AI 爬虫
我只能说…… 打起来。
@samdoehart1333
If only they could BOTH lose. That would be amazing.
要是他们俩能一起输就好了。那可就太爽了。
@asdfpoiuyify
While Meta might escape the “AI training” part of the copyright claim, I don’t think they ‘ll be able argue that hosting torrents is transformative lol.
Meta 或许能逃脱指控中的 “AI 训练” 这部分指控，但他们总不能说托管种子文件是 “转化性使用” 吧，笑了。
@makarambles
As a dev: automated tools are not AI, no matter how big tech companies tell you otherwise to make AI seem more useful. Web scrapers have existed almost as long as the web itself has.
It would be pretty easy to scrape the web for keywords like their media titles + "free download" for example, then have some intern go down the list to see if a site is actually hosting the content.
作为开发者，我想说：自动化工具根本算不上 AI，不管科技巨头怎么吹，说AI有多厉害，绝不是那样。网络爬虫差不多从互联网诞生起就有了。
想收集信息的话其实很简单，比如全网爬取比如某些媒体网站带“免费下载” 之类关键词的内容，然后让实习生一条条核对，验证一下网站是不是真有存有这些内容即可。
@diestormlie
I'm not saying it is obviously Copyright Infringement, but I am saying that Meta was certainly going about it in a clandestine, surreptitious manner. If it was Fair Use and they had every right to do it, why put so much effort into concealing it?
先不说这是不是典型的版权侵权，但 Meta 获取这些内容的方式确实很隐蔽、鬼鬼祟祟。如果这是合理使用，他们完全有权这么做，但他们为什么要费这么大劲藏着掖着呢？
@viper3391
if the are monitoring torrents and sniffing packets, doesn't that mean they themselves are partaking in the distribution network, anyone on a torrent should be technically considered a pirate and liable for infringement.
如果他们在监控种子流量并嗅探数据包，这是不是意味着他们自己也参与了网络分发？按理说，任何参与BT下载的人都应被视为盗版者并承担侵权责任。
@voltairesarmy6702
My guess: They need the adult content to train the system about adult content. They are getting backlash for generative AI creating porn like emma watson porn (real shame how she has been treated). So they disallow it by training it to know what it is and have systems to disallow it.
我猜：他们需要成人内容来训练系统识别成人内容。生成式人工智能生成了像‘艾玛・沃特森色情内容’这类色情作品，这让他们备受抨击（她受到这样的对待实在令人不齿）。所以他们不允许这类内容出现，方法就是训练系统识别什么是色情内容，并建立相应的限制机制。