OpenAI accidentally deleted evidence in copyright case with The New York Times

The New York Times filed a lawsuit against OpenAI and Microsoft at the end of last year for copyright infringement. Then OpenAI responded to the lawsuit filed by The New York Times and said that the media organization manipulated the commands. Finally, OpenAI accused The New York Times of hacking ChatGPT to file a copyright lawsuit.

Now, The New York Times and Daily News have presented an important development to the court in their copyright infringement lawsuit against OpenAI. Lawyers for the publishers claimed that OpenAI engineers accidentally deleted data that could be relevant to the case.

Previously, OpenAI agreed to dedicate two virtual machines so Times and Daily News lawyers could search for copyrighted content in AI training sets. Lawyers for the publishers stated that they have spent more than 150 hours examining OpenAI’s training data with experts since November 1. However, on November 14, it was revealed that OpenAI engineers had deleted the search data stored in one of the virtual machines.

In the written statement submitted to the US District Court for the Southern District of New York, it was stated that an attempt was made to restore the data, but the file names and folder structures were irreversibly lost. This made it impossible to detect how copyrighted content was used in artificial intelligence models.

Publishers’ lawyers stated that they had to start their work from scratch due to lost data and that they spent both serious manpower and computer processing resources in the process. Although there is no evidence that this data loss was intentional, the lawyers emphasized that what happened showed that OpenAI was in the best position to examine its own data sets and detect potential breaches. There was no statement from OpenAI on the subject.