Is It Legal To Train AI Models On Copyrighted Books?
· Technology · TechCrunch
AI models powering chatbots like ChatGPT, Gemini, and Claude are trained on massive databases containing hundreds of millions of books, online articles, and academic papers. Many published authors contributed to these tools without their knowledge or consent, raising questions about legality. In a notable ruling last year, Judge William Alsup ordered Anthropic to pay a $1.5 billion copyright settlement to a group of writers. However, the judge ruled that Anthropic’s AI training itself was lawful, penalizing the company specifically for using pirated books from illegal online shadow libraries. Intellectual property attorney Cathy Gellis notes that copyright law hinges on copying rather than consuming or reading a work, making the ruling broadly advantageous for AI companies. Because copyright legislation has not been updated since 1976, judges are left interpreting 50-year-old guidelines to address complex modern legal questions surrounding artificial intelligence.
Why it matters
The intersection of copyright law and artificial intelligence impacts authors whose livelihoods face disruption and tech companies investing billions in model training. Current legal ambiguity leaves the future of AI development and creator compensation unsettled.
Read the original report — TechCrunch
Join us on Telegram
Breaking news the moment it lands. At 10,000 members we ship the Android app.