Thomson Reuters has won a significant fair use victory in an AI copyright infringement case against Ross Intelligence, potentially setting a precedent for future AI-related copyright disputes.

The case, approaching its fifth anniversary, centers on Ross Intelligence's alleged unauthorized use of Thomson Reuters' Westlaw legal database content to create an AI-powered law search engine. A third-party legal services provider allegedly accessed the protected materials after Thomson Reuters declined Ross's licensing proposal.

Judge Stephanos Bibas ruled that Thomson Reuters' headnotes constitute original, copyrightable works, comparing legal editorial judgment to sculpture: while raw judicial opinions (like marble) aren't copyrightable, the creative selection process produces protected work.

Key Findings:

• Ross's use was deemed non-transformative
• The court found evidence of "actual copying" in 2,243 of 2,830 examined headnotes
• The ruling specifically addresses non-generative AI, not covering broader AI applications
• Ross allegedly aimed to create a market substitute for Westlaw

This victory could influence music industry copyright suits against AI developers, though the ruling's scope is limited to non-generative AI. The judge explicitly noted: "Because the AI landscape is changing rapidly, I note for readers that only non-generative AI is before me today."

The decision challenges AI developers' claims that mass ingestion of copyrighted works for AI training constitutes fair use, though its direct application to generative AI cases remains unclear.

The ruling establishes an important precedent for copyright protection in AI development while leaving open questions about generative AI's broader legal implications, particularly in cases involving creative works like music and literature.

Related Articles

Previous Articles