OpenAI accused of hiding chat logs in NYT copyright battle

OpenAI faces serious sanctions in a high-stakes copyright lawsuit led by The New York Times after allegations that it concealed and improperly handled ChatGPT output logs that could prove or refute infringement. Plaintiffs say OpenAI had large de-identified log samples and misrepresented their accessibility, then used heavy redactions and deletions to hinder discovery, while OpenAI counters that it protects user privacy and fair use. The court constrained access to a 20 million-log sample and highlighted issues with a much larger 78 million-log set, raising the possibility that sanctions could bar the use of the 20 million sample and weaken OpenAI’s defense in a case that could hinge on whether training on copyrighted content constitutes fair use.