Federal court rules AI startup infringed on Thomson Reuters’ copyright

Thomson Reuters filed the lawsuit in 2020, alleging that Ross Intelligence had improperly reproduced material from its legal research platform, Westlaw. Last week, US District Judge Stephanos Bibas ruled in favor of Thomson Reuters, finding that Ross had infringed on the company’s copyright.

“None of Ross’s possible defenses holds water. I reject them all,” Bibas wrote in his decision.

The case is part of a growing legal debate over how AI companies use copyrighted material to train their models. Many AI tools are developed by training on copyrighted works, including books, films, and websites. Wired reported that there are currently several dozen similar lawsuits in the US, with additional legal challenges in countries such as China, Canada, and the UK. Ross Intelligence, which shut down in 2021 due to litigation costs, is among several AI firms facing lawsuits over copyright issues.

Bibas ruled in favor of Thomson Reuters regarding its argument of fair use. Fair use allows limited use of copyrighted works without permission in certain cases, such as parody, news reporting, or research. Courts consider four factors in determining fair use including the purpose of the use, the nature of the original work, the amount used, and the impact on the market value of the copyrighted material. Thomson Reuters prevailed on two of these factors but Bibas emphasized the fourth, ruling that Ross “meant to compete with Westlaw by developing a market substitute.”

James Grimmelmann, a professor of digital and internet law at Cornell University, described the ruling as a major setback for AI companies.

“If this decision is followed elsewhere, it's really bad for the generative AI companies,” Grimmelmann said, according to Wired. He also explained that Bibas’ decision suggests much of the legal precedent AI companies have relied on for fair use may be “irrelevant."