Your training data just became a personal liability. Sony Music Publishing and Warner Chappell sued Anthropic — and named Dario Amodei and Ben Mann as individual defendants. Tens of thousands of songs, up to $150K each. "We just scraped it" stopped being an answer in 2026. If you fine-tune or run RAG, write down where every dataset came from before someone subpoenas it. Provenance is cheap now, ruinous later. More at www.contentbuffer.com Honestly — can you name the source of every dataset in your stack? #AI #Copyright #DataGovernance #MachineLearning #LegalTech #TechNews