In a major victory for copyright holders, Thomson Reuters has won the first significant AI-related copyright case in the United States.
The legal dispute, which began in 2020, involved claims that Ross Intelligence, an AI-driven legal research startup, had unlawfully copied materials from Westlaw, Thomson Reuters’ legal research platform.
A judge ruled in favor of Thomson Reuters, dealing a serious blow to AI companies that rely on copyrighted materials to train their models.
Court Rules Ross Intelligence Violated Copyright Law
The case was decided by US Circuit Court Judge Stephanos Bibas, who delivered a strong rebuke against Ross Intelligence. In his ruling, he dismissed all of the AI company’s defenses, stating:
“None of Ross’s possible defenses holds water. I reject them all.”
Ross Intelligence, which had already shut down operations in 2021 due to the financial burden of litigation, did not respond to requests for comment. However, Thomson Reuters welcomed the decision.
Jeffrey McCoy, a spokesperson for the company, emphasized the importance of the ruling:
“We are pleased that the court granted summary judgment in our favor and concluded that Westlaw’s editorial content, created and maintained by our attorney editors, is protected by copyright and cannot be used without our consent. The copying of our content was not ‘fair use.’”
AI Copyright Battles Are Just Beginning
This case is just one of many legal challenges AI companies are facing over the use of copyrighted material. The rise of generative AI, which often relies on books, films, artwork, and other copyrighted content for training, has sparked a wave of lawsuits.
Similar legal battles are unfolding across the United States, UK, Canada, and China. Major AI players, including OpenAI and Google, are currently facing multiple copyright lawsuits, raising questions about how AI companies can legally train their models.
Fair Use Argument Falls Apart in Court
One of the biggest takeaways from this ruling is how fair use, a key defense for AI companies, was rejected by the court. The fair use doctrine allows limited use of copyrighted materials without permission in certain cases, such as parody, research, or news reporting.
Courts consider four factors when determining fair use:
- Purpose and character of the use
- Nature of the copyrighted work
- Amount of the work used
- Effect on the market value of the original work
While Thomson Reuters won on two of these four factors, Judge Bibas placed the greatest emphasis on the fourth, ruling that Ross Intelligence was creating a direct market competitor to Westlaw, which tipped the scales against fair use.
What This Means for AI Companies
This decision could have major implications for AI firms that rely on copyrighted data. James Grimmelmann, a professor at Cornell University specializing in internet law, warned that this ruling might be a major setback for the generative AI industry:
“If this decision is followed elsewhere, it’s really bad for generative AI companies.”
Legal expert Chris Mammen, a partner at Womble Bond Dickinson, agrees. He believes that while the ruling doesn’t eliminate fair use as a defense entirely, it makes it much harder for AI companies to rely on it.
“It puts a finger on the scale towards holding that fair use doesn’t apply.”
The Future of AI and Copyright Law
With this ruling, the legal landscape for AI firms is becoming increasingly challenging. While tech giants like OpenAI and Google have the financial resources to fight lengthy legal battles, smaller startups could struggle to survive these lawsuits.
As courts continue to weigh in on these issues, one thing is clear: the relationship between AI and copyright law is far from settled. This case could set an important precedent for future legal disputes over how AI tools are trained and what constitutes fair use.
