Thomson Reuters Wins Legal Battle Over AI and Copyright
Thomson Reuters has won an important legal battle that could have major consequences for the AI industry. A court has ruled that Ross Intelligence infringed copyright by using Westlaw's content without permission. What does this mean for the future of AI and copyright?
Thomson Reuters has won an important legal victory in one of the first lawsuits to test how AI is allowed to use copyrighted material. A judge in the US District Court in Delaware has ruled that the legal AI company Ross Intelligence infringed Thomson Reuters' copyright by using content from their Westlaw database without permission.
Early Ruling Could Have Major Consequences for the AI Industry
The case, which began in 2020, could have far-reaching consequences for how AI models are trained and whether they can claim "fair use" when using copyrighted material. Similar lawsuits are pending against AI giants like OpenAI and Microsoft, meaning this ruling could serve as an important guide for future decisions.
The Court: No Doubt the Content Was Copied
In his ruling, Judge Stephanos Bibas dismissed Ross Intelligence's defense and found that their AI tool had used copyrighted material in a way that can't be classified as fair use. He especially emphasized that Ross had created a direct competitor to Westlaw by using its material to train its AI.
Thomson Reuters commented on the ruling in a statement to The Verge:
"We're pleased the court confirmed that Westlaw's editorial content, created by our legal editors, is protected by copyright and can't be used without our permission. The copying of our content was not 'fair use'."
Ross Used Westlaw's Data Through a Third Party
Westlaw contains both non-copyrighted information, such as case law, and original content, such as summary "headnotes" written by lawyers. These headnotes are an important feature of Westlaw and one reason the service is so valuable to lawyers.
According to the ruling, Ross Intelligence first tried to license Westlaw data directly from Thomson Reuters but was rejected. The company then turned to a third party, LegalEase, and bought 25,000 "Bulk Memos" - legal questions and answers based on Westlaw's headnotes - to use in AI training.
Ross CEO Andrew Arruda has claimed that Westlaw data only made up "noise" in the AI, and that the tool actually focused on extracting answers directly from legal texts. The court dismissed this argument, finding the copying was so obvious that "no reasonable jury could reach a different conclusion".
What Does This Mean for the Future?
The ruling against Ross Intelligence shows that companies building AI tools can't freely use copyrighted material without facing legal consequences. Since more similar lawsuits are pending, this could become a precedent affecting the entire industry.
How should AI companies handle copyright in the future? Will this lead to stricter regulations around AI training? The discussion is far from over.
Source: The Verge