Business & Strategy · intermediate · concept 310 of 311 · Updated Nov 4, 2026
AI, Copyright and Training Data
AI and copyright asks whether training a model on copyrighted works without permission is lawful, and whether AI output can be copyrighted. It matters because the answers decide what data a company may use and what it owns. The US Copyright Office holds that purely AI-generated material cannot be registered. In June 2025 a US judge ruled that training Anthropic's models on bought books was fair use, but keeping pirated copies was not.
Top AI contributors on this
- Sam AltmanOpenAI
- Fei-Fei LiWorld Labs / Stanford University
- Kai-Fu LeeSinovation Ventures / 01.AI
- Timnit GebruDistributed AI Research Institute (DAIR)
- Margaret MitchellHugging Face
- Elon MuskTesla / SpaceX
Key terms
Learn these first
Where you meet it in the real world
Data licensing between publishers and AI labs, registering works with AI-assisted parts, vendor contract indemnities, training-data audits, EU AI Act compliance for model providers
Videos
Simons Institute for the Theory of Computing · YouTube
Schwartz Reisman Institute · YouTube
CITRIS and the Banatao Institute · YouTube
Guides and articles
Courses, papers, and more