OpenAI claims that it is “impossible” to train an AI model without copyrighted material

This is supposed to be a proactive defense of OpenAI against the many accusations of copyright infringement, but not everyone suing the company will see it that way.
Source link