NYT alleges OpenAI concealed copyrighted content evidence in trial

The New York Times and The Daily News have escalated their two-year copyright lawsuit against OpenAI, alleging the company hid evidence about its ability to search ChatGPT chat logs and training datasets for copyrighted works. Throughout the case, OpenAI argued it lacked the technical capacity to search its massive training corpus and that retrieving logs would be burdensome and raise privacy concerns. However, an April court-ordered deposition from OpenAI data privacy engineer Vinnie Monaco allegedly revealed that OpenAI had already conducted internal searches of its training data for copyrighted journalism and had compiled a database of about 78 million de-identified ChatGPT conversations. Monaco also disclosed that shortly after the lawsuit was filed, OpenAI implemented a 'Bloom' filter as part of 'Project Giraffe' — a toolset designed to detect and record regurgitation of copyrighted content in outputs.

These revelations directly contradict OpenAI's prior claims of technical infeasibility. The plaintiffs had requested a sample of 120 million chat logs, which OpenAI negotiated down to 20 million. When finally produced in December, the sample was so heavily redacted that the court deemed it 'unusable.' The plaintiffs also accuse OpenAI of deleting billions of ChatGPT outputs after the lawsuit was filed, violating a court preservation order, and substituting millions of logs in the requested sample. They are now asking the judge to discipline OpenAI: prevent use of the unreliable sample, accept that logs would show major regurgitation, bar OpenAI from arguing otherwise, and force payment of legal fees. OpenAI spokesperson Drew Pusateri denied the allegations, accusing the Times of trying to invade user privacy as its case weakens, and reaffirmed OpenAI's commitment to fair use doctrine.