Normalizing AI as being human equivalent means the AI is legally culpable for its own actions rather than its creators or the people using it, and not guilty of copyright infringement for having been trained on proprietary data without consent.

At least I think that's the plan.