> I think that the de facto situation is that AI does not have to know about licences or copyright at all.
That is simply not true.
Companies can get into legal trouble if they don't.
Copilot does that bookeeping:
https://docs.github.com/en/copilot/how-tos/get-code-suggesti...
> Companies can get into legal trouble if they don't.
Heard of Meta torrenting copyrighted material? What kind of trouble did they get into?