I believe it when I see it!

This isn’t a particularly useful solution to the problem. The problem is that what the AI companies are doing isn’t against current copyright laws, it’s just a bit unethical.

We need to decide that this should either be against the law or not and then adjust the laws accordingly.

Just saying that they train their model with works protected by copyright is a non-statement, since it’s not a violation of copyright to look at and process someone else’s work.