Topic
NewsCreative IndustriesDevelopingNew
A federal magistrate judge in California has ordered Meta to identify everyone who decided to obtain shadow-library book datasets for its Llama models; it is a discovery order, and no court has found Meta liable for the downloads.
Evidence: 3 sources · 3 independent reports · 3 open questions
NewsCreative IndustriesNew
The Third Circuit upheld a ruling that ROSS Intelligence infringed Thomson Reuters' copyrights by training a legal research tool on Westlaw headnotes, the first US appellate fair-use ruling on AI training, though the court said it was not deciding questions about generative AI.
Evidence: 5 sources · 2 official records · 3 independent reports · 3 open questions
ExplainerCreative IndustriesNew
Disputes over AI and creative work turn on a few separate questions about training data, generated output, and who can own it, and the answers differ by country.
Evidence: 1 source · 1 official record