TL;DR
The Third Circuit has upheld the first U.S. appellate ruling that using copyrighted legal headnotes to train a competing AI tool is not fair use, setting a precedent that could reshape how AI companies license training data.
What happened
- The US Court of Appeals for the Third Circuit issued its opinion on September 29, 2026, affirming a lower court ruling against Ross Intelligence.
- Thomson Reuters (owner of Westlaw) sued Ross Intelligence for using Westlaw headnotes, short legal summaries, to train a competing AI-powered legal research tool.
- A three-judge panel rejected Ross's argument that transforming the summaries into AI training data constituted fair use.
- The opinion was initially filed as a sealed document, signaling sensitivity around the ruling's scope.
- This is the first U.S. appellate decision to directly address whether AI training on copyrighted material qualifies as fair use.
Why it matters
- Fair use as an AI training defense is now weaker: the court found that using copyrighted content to build a competing product does not clear the transformation bar, even as an intermediate step.
- Every AI company training on proprietary data faces heightened legal exposure; the ruling gives copyright holders a concrete appellate precedent to cite.
- Licensing costs for training data could rise sharply as content owners gain leverage to demand payment before AI developers can legally use their material.
- The decision lands in the legal tech sector first, but the logic applies broadly to any domain where AI tools compete with the original content source.
- Ross Intelligence's core argument, that AI training is sufficiently transformative to qualify as fair use, was explicitly rejected at both the district and appellate levels.
What to watch next
- Whether Ross Intelligence petitions for en banc review or seeks Supreme Court certiorari, which would determine if this precedent holds or gets revisited.
- How other AI training copyright cases (music, news, books) cite or distinguish this ruling in their own fair use arguments.
- Whether major AI labs accelerate licensing deals with data providers now that the appellate shield of fair use has been weakened in a key circuit.
Originally published on Present of AI, a daily source-linked AI news timeline. Read the full timeline or browse the open dataset.