This is a contentious statement - the large datasets used to train all the new generative AI projects are comprised, at an individual level, of "other people's stuff".

Copyright, trademark, IP in general, there's a 9.0 magnitude earthquake coming.