- Thomson Reuters won a first victory in a copyright affair linked to AI against Ross Intelligence.
- The decision highlights the limits of fair use. The fair use is at the center of prosecution under adaptation in matters of copyright of the AI.
- Legal experts claim that the case differs from other disputes involving AI generating companies.
Technological content and conglomerate Thomson Reuters This week marked the first big victory in a case of copyright linked to American artificial intelligence.
The decision of the Delaware of a federal judge in favor of Thomson Reuters on the legal doctrine of “fair use”, however, does not mean that the series of authors and publishers who have pursued companies of generators generators As Openai for copyright violation can expect to have the same type of success, technology law experts, Business Insider, told experts.
The decision, however, could have an influence on the outcome of the cases which are currently taking place before the courts, said one of the experts in technology law.
Thomson Reuters continued the legal startup of the AI now closed Ross Intelligence in 2020, arguing that Ross broke its copyright when the startup used the content of Thomson Reuters’ Westlaw Legal research database to create a competing platform that uses artificial intelligence.
In his decision revised on Tuesday, the judge of the American district court Stephanos Bibas killed the defense of Ross for fair use and rather rendered a summary judgment for Thomas Reuters on fair use.
“None of Ross’ possible defenses contains water. I reject them all,” Bibas wrote in the decision, in which he explained that, in his opinion in 2023 in the case, he denied a summary judgment on fair use.
The question of fair use is at the heart of the main lawsuits on copyright against AI AI companies. Companies like Openai have indicated the exemption of exemption on the fair use of copyright laws in order to justify their use of material protected by copyright to form AI models.
Legal experts told Bi that there were key nuances in the Thomson Reuters file against Ross and the successful dispute involving companies generating AI – one being that the Thomson Reuters affair did not To do with generative AI technology.
Bibas made this distinction in his decision, writing: “Ross used Thomson Reuters ‘head notes as data on AI to create a legal research tool to compete with Westlaw. It is indisputable that Ross N’ AI is not a generative AI (AI that writes new content itself).
Ross AI was not a generative AI, said the judge. Qi Yang / Getty images
Mark Bartholomew, professor of law at the University of Buffalo, told Bi that he did not think that the decision will have a dramatic effect on other important copyright affairs linked to AI, because it is is that an opinion of a lower jurisdiction and “neglects some of the most important cases of just use.”
“However, the complainants of these coming affairs, like the New York Times, must be satisfied with the result here,” said Bartholomew, adding that these complainants “will probably deceive this decision in their memories.”
The New York Times continued Openai for copyright violation in 2023. The New York Times refused to comment on this story. OPENAI did not immediately respond to a request for comments.
One of the other ways in which the Thomson Reuters affair is different from other disputes during copyright violation linked to AI is that Ross Intelligence was a direct competitor of Thomson Reuters, Bartholomew said.
“In many of the other cases of copyright to decide, the defendant is accused of having copied to form the AI to a goal that does not compete directly with the original,” said Bartholomew . “This can allow this case to be distinguished, allowing AI platforms to assert that they always satisfy the defense of fair use regardless of participation in Thomson.”
Harry Sourden, professor at the Faculty of Law of the University of Colorado, said that he feared that judges to supervise other AI law actions would not take into account the differences between business .
The decision is “likely to confuse judges and others on the question of generative AI, but there are important differences which are very subtle and most people are missing,” said Sourden, “it is times a different technology and the company acted as kind of duplication here. “
For these reasons, Sourden said that the decision should not have much influence on copyright cases in the current AI. He thinks it will still be.
“The judges tend not to be experts in copyright law or generative AI, so I think they will probably miss many of these nuances, and I think that will unfortunately have an effect on other cases “Said Sourden.
James Gatto, partner of the law firm Sheppard Mullin, who co -edit the IA industry team, told Bi that it remains to be seen what type of training effect, if applicable , the decision in favor of Thomson Reuters would have.
“Fair use is a specific survey of single facts in each case,” said Gatto.
Although the judge’s decision does not deal with the generative AI, Gatto said that it “strengthens the limits of fair use, in particular in cases where the material protected by copyright is used to non -transformative purposes to develop a competitive product “.
“The court easily determined that the use of Ross was commercial, because Ross sought to take advantage of the material protected by copyright without paying the usual price,” said Gatto.
Meanwhile, Thomson Reuters welcomed the legal victory, saying in a declaration to Bi: “We are happy that the court made a summary judgment in our favor and concluded that the editorial content of Westlaw created and maintained by our editors in Chief, is protected by copyright and cannot be used without our consent.
Ross representatives did not immediately respond to a request for comments.