Open AI Copyright vs. AI: The $1.5B Turning Point
5 snips
Sep 15, 2025 A landmark $1.5 billion settlement between an AI company and writers could reshape the landscape of copyright law. The discussion highlights the court's ruling on the legality of using pirated data for AI training and the financial implications for the involved company. It also dives into the complexities of copyright challenges in AI image generation, revealing how companies like Adobe grapple with crediting original creators. This could define future interactions between copyright holders and AI industries.
AI Snips
Chapters
Transcript
Episode notes
Historic Settlement Sets Legal Tone
- The Anthropic settlement is the largest payout in U.S. copyright history and may set a major legal precedent.
- The deal signals courts may accept training on purchased works as transformative fair use while penalizing piracy sources.
Shadow Libraries Boosted Claude's Tone
- Jaeden explains Anthropic used pirated "shadow libraries" of books to rapidly improve Claude's writing tone.
- He recounts that Anthropic later bought many books and scanned them to include in training as well.
Judge Ruled Training Can Be Fair Use
- Judge William Alsup ruled training on copyrighted material can be transformative and protected by fair use.
- The court differentiated legal training from illegal piracy, allowing purchased-content training but condemning pirated sources.