Yes lol. This is the same thing as artists being mad that they put their images on the internet and then people saw those images on the internet.

If your content is posted publicly I don't see how you can be mad that someone put it into an algorithm to turn it into a set of vector embeddings determining which words are likely to come after the previous ones (or embedding "feature" information about an image, same thing).

Yes and no. There is a file called, "robots.txt", which sets crawling limits for the site. Nothing stops crawlers from crawling past (unless account restrictions exist), but it also sets a legal standard. Many sites' contents are crawled or indexed because of this defacto standard.

That said, copyright, which is the actual claim here, is pretty cut and dry. The AI is digesting the copyrighted contents to form at least part of its language model. This legally means the language model is a derivative work, which means the AI is in violation of copyright laws.