OpenAI and NYT lawsuit: logs, evidence and the copyright dispute

New York Times and The Daily News have accused OpenAI of withholding evidence in a copyright case.

According to their filing, the company denied that it could search chat logs and training data, even though such checks were already being carried out internally.

The plaintiffs also allege that OpenAI had a database of roughly 78 million anonymised conversations and tools to detect repeated outputs.

They are now asking the court to restrict the use of part of the logs and to sanction OpenAI for alleged discovery violations.

Why it matters: this dispute highlights how sharply copyright, transparency and data protection intersect in AI.

Where do you draw the line between user protection and evidence disclosure?