Archive — US — techcrunch
New York Times says OpenAI hid evidence in ChatGPT copyright trial
The New York Times and The Daily News accuse OpenAI of withholding evidence in a copyright trial concerning the use of their journalism in ChatGPT outputs. Allegations include OpenAI falsely claiming it couldn't search its chat logs, while internal records suggest otherwise. They seek judicial action against OpenAI for allegedly hiding information and improper evidence management in court.
Key points
- The NYT and Daily News claim OpenAI lied about its ability to search for copyrighted content in training data.
- OpenAI argues searching and processing chat log data would violate user privacy.
- A deposition revealed OpenAI had conducted internal searches prior to the lawsuit, contradicting its claims.
- OpenAI reportedly built a database of 78 million de-identified ChatGPT conversations related to copyright evaluation.
- The plaintiffs argue OpenAI's provided chat logs were heavily redacted and unreliable.
- They are asking the court to discipline OpenAI for allegedly obstructing the discovery process.
Why it matters
- This case reflects ongoing debates over copyright law in the AI landscape amidst an increasing reliance on generative AI tools.
- Similar lawsuits and controversies highlight the struggle for content creators in the digital age.
- As of 2025, AI trends include discussions on transparency in AI training data and user privacy, recently spotlighted by laws and guidelines shaping AI usage.
The NYT and Daily News's allegations against OpenAI underscore the significant challenges in navigating copyright law in AI, stressing the need for clearer standards and accountability as the use of AI-generated content expands.