Major News Outlets Seek Sanctions Against OpenAI Over Copyright Evidence
Key Takeaways
- ▸Newspapers are seeking court sanctions against OpenAI for allegedly hiding evidence about how ChatGPT was trained on copyrighted news content
- ▸OpenAI's claims that privacy concerns prevent log disclosure are being challenged as pretextual obstruction of discovery
- ▸The "fair use" defense for training AI on copyrighted material is being tested across multiple industries and jurisdictions with uncertain outcomes
Summary
The New York Times, the Daily News, and other major news outlets have filed a motion in federal court seeking sanctions against OpenAI for allegedly hiding evidence crucial to a landmark copyright infringement lawsuit. The newspapers claim OpenAI is engaging in "discovery misconduct" by refusing to release datasets and ChatGPT logs that would demonstrate how the AI system was trained on millions of copyrighted news articles without permission or compensation.
The filing accuses OpenAI of making "misrepresentations" for two years about its technical ability to identify and retrieve copyrighted content from its training datasets. Recent depositions of OpenAI employees have contradicted the company's earlier claims about these limitations. OpenAI counters that restrictions on sharing ChatGPT logs are necessary to protect user privacy, and maintains that AI training on published internet content falls under the "fair use" doctrine of U.S. copyright law.
The case is part of a broader legal reckoning over AI copyright practices, with novelists, visual artists, and musicians also pursuing lawsuits against AI companies with mixed results. Most significantly, OpenAI rival Anthropic recently agreed to pay book authors $1.5 billion for training its Claude chatbot on pirated works—a settlement that suggests courts may find AI companies liable for copyright infringement despite fair-use arguments. The outcome could reshape how AI systems are trained and threaten the already-struggling news industry's viability.
- Anthropic's $1.5 billion copyright settlement with book authors signals that courts may find AI companies liable for using copyrighted training data
