I think this completely misses the point... the point is that Meta pirated the media they used to train their model.

I am not a fan of US copyright law, but if I torrented millions of books, I would be facing a felony charge in criminal court and a (with statutory damages as high as $150,000 per title in cases of willful infringement) multi-billion dollar lawsuit in civil court.

In my opinion, this has nothing to do with whether or not AI training is transformative and this fair use, and everything to do with whether or not the laws apply to everyone equally. If Facebook isn't forced to pay billions and elect a sacrificial executive to serve prison time, then I will remain angry.