Publishers File Brief in Response to Internet Archive Appeal
On Friday, the publishers in Hachette, et al. v. Internet Archive filed their response brief opposing the archive’s appeal of its loss in district court. IA maintains that its practice of “Controlled Digital Lending” is not copyright infringement under the doctrine of fair use despite the lower court’s fast and resounding decision rejecting that defense in late March 2023. As the lower court stated:
At bottom, IA’s fair use defense rests on the notion that lawfully acquiring a copyrighted print book entitles the recipient to make an unauthorized copy and distribute it in place of the print book, so long as it does not simultaneously lend the print book. But no case or legal principle supports that notion. Every authority points the other direction.
Given the amount of Second Circuit precedent contributing to the district court’s four-day turnaround decision, it is hard to see how IA will fare any better on appeal. But we shall see.
Santos v. Kimmel May Be Entertaining
In a complaint filed in mid-February, American fabulist George Santos accuses comedian Jimmy Kimmel et al. of copyright infringement, fraudulent inducement, breach of contract, and unjust enrichment, all arising from Kimmel’s pranking Santos’s new gig making personalized video messages on Cameo.com. Kimmel allegedly tricked Santos by creating fake customer identities and then, posing as those “customers,” requested absurd material for Santos to record. Kimmel then played five of the Cameo videos on his TV show to mock Santos.
I read the Santos complaint over the weekend, and if the facts presented are accurate, the case could provide some interesting details for copyright nerds. But given that we’re talking about George Santos, the prudent course at the moment is to at least wait for the defendants’ response before commenting.
Goldsmith and Andy Warhol Foundation Settle
In a court filing on Friday, photographer Lynn Goldsmith and Andy Warhol Foundation (AWF) agreed to settlement terms, concluding the landmark fair use case. AWF will pay Goldsmith $10,250 based on the original licensing fee for use of her photo in the Warhol screen prepared for the Conde Nast issue, and it will pay another $11,272.94 in taxable costs. Both parties are responsible for other expenses and attorney fees.
“AWF’s position is that the original creation of the Prince Series was fair use, and that nothing in the Supreme Court’s opinion undermines that view,” the court filing states. Indeed, the legality of the entire Prince Series prepared by Warhol was not addressed in this case. But the clarification on “transformative” use delivered by the Court—and which led directly to judgments against appropriation artist Richard Prince—suggests that AWF should probably not evangelize that opinion in the art world.
More Suing of Generative AI
On March 8th, authors Abdi Nazemian, Brian Keene, and Stewart O’Nan filed a class-action lawsuit against NVIDIA Corporation, alleging copyright infringement of books for the purpose of training the NeMo Megatron–GPT, a Large Language Model (LLM). The claim rests entirely on the allegation of unauthorized reproduction in the training process, and as with the Chabon lawsuit against Open AI, plaintiffs here accuse the defendant of accessing large volumes of books from dubious sources…
Bibliotik is one of a number of notorious “shadow library” websites that also includes Library Genesis (aka LibGen), Z-Library (aka B-ok), Sci-Hub, and Anna’s Archive. These shadow libraries have long been of interest to the AI-training community because they host and distribute vast quantities of unlicensed copyrighted material. For that reason, these shadow libraries also violate the U.S. Copyright Act.
Given the general consistency in both the legal and factual allegations in most of the cases against generative AI developers, the earliest outcomes could signal a blowout for either creators or the developers. As stated in earlier posts, if the reproduction right is held to be violated in the process of machine learning, it is hard to see how any of the developers overcome that claim. The case to watch is arguably New York Times v. Open AI, because the Times has presented so much compelling evidence that works output by the system are substantially similar to works input by the system. In a close second, is probably Concord v. Anthropic, where music publishers have likewise presented evidence of substantially similar lyrics output by the system.
