OpenAI accidentally deleted potential evidence in NY Times copyright lawsuit (updated)

OpenAI allegedly deleted data related to a NY Times copyright lawsuit.

: OpenAI engineers reportedly deleted data relevant to a lawsuit with The NY Times and Daily News. The deletion occurred during their search for copyrighted content in AI training data. OpenAI claims the data loss was due to a system misconfiguration requested by the plaintiffs, not intentional. Licensing deals with publishers like Axel Springer and Dotdash Meredith are not publicly disclosed by OpenAI.

Lawyers for The New York Times and Daily News claim OpenAI engineers accidentally deleted crucial data, impacting their lawsuit against OpenAI for allegedly using copyrighted works in AI training without permission. This occurred after OpenAI provided virtual machines for their copyright content search. OpenAI managed to recover some data, but folder structures and file names were lost, making it challenging to ascertain where the publishers' content was used.

The plaintiffs had to redo searches, wasting time and resources. They acknowledged there was no evidence of deliberate deletion but stressed OpenAI's responsibility in searching its own datasets effectively. An OpenAI spokesperson did not comment on these allegations, leading to further complications in the ongoing lawsuit.

OpenAI's attorneys responded, insisting plaintiffs requested changes that caused the technical issue. Although OpenAI continues to argue for fair use of publicly available data in their models, they have signed several licensing deals, the specifics of which remain private. These deals include paying Dotdash a reported $16 million annually, although OpenAI hasn't specified which copyrighted works might have been used without permission.

Copyright lawsuitData deletionAI Models