[Shrezzing]

Copyright doesn't stop the collection of content, it stops the copying, processing, & redistribution of content. Internet Archive acts as a library, so its widely accepted as fair-use when it makes collections of webpages available.

OpenAI's distribution is materially different to that of a library, so it's not a like-for-like comparison.

One of the main tests of copyright law (at least in the US) is if the entity distributing is _selling_ the copied/derivative work. It's unambiguous that OpenAI is selling something akin to derivative works, which is why NYT feels they can go after this claim. Meanwhile IA's operations don't create sales or incur profits, so while NYT's legal team may be able to establish that copies have been distributed, without the _sale_ aspect of the infringement, judges aren't guaranteed to side with NYT in an legally expensive PR nightmare.