In what context did OpenAI say they aren’t using copyrighted data in the first place?
OpenAI has publicly said on multiple occasions [0] that it does use cooyrighted data for training, and that that use is fair use under the law. They may have said that they don’t use copyrighted data in some narrower context, but I am unfamiliar with the claim.
I also wouldn’t take Elon at his word about anything related to this, given his history of both publicly lying about things he is in a position to know about when it suits him and his history of making up strong statements where he isn't in a position to know, and the fact that he has a commercial interest as someone trying to stand up an OpenAI competitor to spread lies about both what OpenAI has claimed and what OpenAI is factually doing.
OpenAI has publicly said on multiple occasions [0] that it does use cooyrighted data for training, and that that use is fair use under the law. They may have said that they don’t use copyrighted data in some narrower context, but I am unfamiliar with the claim.
I also wouldn’t take Elon at his word about anything related to this, given his history of both publicly lying about things he is in a position to know about when it suits him and his history of making up strong statements where he isn't in a position to know, and the fact that he has a commercial interest as someone trying to stand up an OpenAI competitor to spread lies about both what OpenAI has claimed and what OpenAI is factually doing.
[0] e.g., https://www.uspto.gov/sites/default/files/documents/OpenAI_R...