I would love to see that legal argument given their view of “fair use” of all the copyrighted material that went into OpenAI models.