The defenders of the idea that the formation of the AI is transformer always see Chhabria’s decision as a victory. “Judge Chhabria judged today, in the end, that the training of generative models of AI on the material protected by copyright is clearly transformer, and in the absence of proven damage to the market is fair use,” explains Adam Eisgrau, principal director of AI, creativity and copyright policy in the technological chamber of progress. “He did not like to arrive at this conclusion for reasons which he details and which, with regard to market damages, are completely out of step with a previous fair precedent. The dilution of the market is Malarkey.”
And that’s the catch. Chhabria has taken the trouble to emphasize that its decision was based on the specific set of facts in this case-leaving the door open to other authors to continue Meta for the violation of copyright in the future: “In many circumstances, it will be illegal to copy works protected by copyright to form generative AI models without permission,” he wrote. “This means that companies, to avoid responsibility for copyright violation, will generally have to pay copyright holders for the right to use their materials.”
“On the surface, it looks like a victory for the AI industry,” explains Matthew Sag, professor of law and artificial intelligence at Emory University, noting that Meta clearly won a victory with the recognition of Chhabria according to which the training of AI models is a transformative. “However, The Court Does Take Very Seriously The Idea That Ai Models Trained On Plaintiffs’ Books Could ‘Flood the Market With Endless Amants of Images, Songs, Articles, Books, and More,’ Thereby Harming The Market For The Original Works. He probably Takes It More Seriously Than Given that they did not put any evidence on this outcome.
“The court ruled that the companies of AI which` fuel the copyright work in their models without obtaining the permission of copyright holders or paying them “generally violated the law”, said the lawyers of the plaintiffs of Boies Schiller Flexner in a press release. “However, despite the undisputed unprecedented hacks of Meta, the works protected by copyright, of Meta. We are respectfully in disagreement with this conclusion. “
The Meta team had a sunnier response. “We appreciate today’s decision of the Court,” said Meta spokesman Thomas Richards in a statement. “Open-Source AI models feed transformative innovations, productivity and creativity for individuals and businesses, and the fair use of copyright equipment is a vital legal framework to build this transformative technology.”
Applicants in other cases of AI pay particular attention to the result. “We are disappointed with the decision, but only in part,” explains Mary Rasenberger, CEO of the author’s guild, who continues Openai in her own copyright violation case, noting that Chhabria has kept the decision deliberately closely.
“In the big scheme of things, the consequences of this decision are limited. It is not a collective recourse, so the decision affects only the rights of these 13 authors – not the innumerable others whose meta works used to form its models,” wrote Chhabria. “And, as should now be clear, this decision does not represent the proposal that Meta use of documents protected by copyright to form its linguistic models is legal.”
