Federal judge sides with Meta in lawsuit over training AI models on copyrighted books

On Wednesday, a federal judge supported Meta in a case against the author of six books, including Sara Silverman, who complained that the company had illegally trained its AI models in their copyright work.

Federal judge has issued Vince Churia Short verdict – That means the judge was able to decide on the case without sending any jury – in favor of Meta, the training of the AI ​​models in the copyright books in this case was under the “fair use” of copyright law and thus legal.

Comes just a few days after the decision Federal Judge In similar cases. Together, these cases are a win for the technology industry, which has spent years of legal fighting with media agencies that argue that copyrighted work is the only use of AI models in training.

However, these decisions are not to win for some of the expected companies – both judges mentioned that their cases were limited to opportunities.

The judge made clear that this decision does not mean that all AI model training in copyrighted works is legal, but in this case the plaintiffs have “made the wrong argument” and fail to develop adequate evidence in support of the correct.

Judge Chambariya said in his decision, “The verdict does not stand in favor of the proposal that the use of copyrighted materials for training its language models is legal.” Later, he said, “The plaintiffs will often win, at least where these cases have a better and better record on the market influence of the dispute,” he said.

Judge Chambariya ruled that in this case the meta was converter to the use of copyrighted works – meaning the company’s AI models did not just reproduce the authors’ books.

Furthermore, the plaintiffs failed to convince the judge that copied the Meta books damaged the market for those authors, which is the main reason for determining whether copyright law has been violated.

“The plaintiffs have not presented any meaningful proof of the market decrease,” the judge said.

AI models are involved in the books of both ethnographic and meta winners, but there are several active cases against technology companies for training AI models in other copyrighted works. For example, The New York Times is suing against Openai and Microsoft AI models in news articles for training, when Disney and Universal Mid Journey is suing against For training of AI models on film and TV shows.

Judge Chambariya mentions in his decision that the defense of a fair use depends a lot on the details of a case and the logic of fair use may be stronger than others in some industries.

“Looks like a certain type of job markets (such as news articles) may be more risky in indirect competition than AI outputs,” said Churia.