The latest AI‑related litigation wave includes a major copyright case filed by USA Today and its 13 sister papers against OpenAI. Filed on Oct. 8, 2026 in Manhattan federal court, the suit alleges OpenAI copied "hundreds of thousands" of articles—over 160,000 WebText entries, including about 83,300 USA Today pieces—to train its GPT models without permission. The publishers are seeking more than $250 million in damages, an injunction to stop further infringement, and destruction of any models built on the disputed data. This action follows a string of high‑profile lawsuits against OpenAI by other news outlets such as The New York Times, Encyclopaedia Britannica and a coalition of 400 local newspapers, signaling a coordinated push by the media industry to protect copyrighted content from AI training.
A recent appellate decision also shapes the legal landscape: the Third Circuit affirmed that Ross Intelligence’s use of Westlaw headnotes for its AI legal‑research tool was not fair use, reinforcing that copyrighted “creative spark” material used for AI training can fall outside the fair‑use defense. While the Ross case involved a non‑generative tool, its reasoning is likely to be cited by publishers as they argue that large‑language models like ChatGPT similarly benefit from copyrighted works.
Why it matters: These lawsuits could force AI firms to obtain licenses for training data, alter how large language models are built, and potentially limit the breadth of publicly available AI services. Courts’ interpretations of fair use in the AI context will set precedents that could either curb or permit expansive data scraping.
What to watch next:
1. Upcoming hearings in the Concord Music Group v. Anthropic case, which will test copyright limits on song lyric training data.
2. The New York Times v. OpenAI and Microsoft MDL, where a summary‑judgment on fair use for large‑language models is pending.
3. Potential Supreme Court review of AI‑training fair‑use disputes, especially if lower‑court rulings diverge.
4. Any settlement deals where AI companies negotiate bulk licensing with news publishers, which could reshape industry standards for AI data usage. [[1, 2]]