We didn't actually pirate this book.

But according to Judge Chhabria and Meta Platforms Inc., you could have, and it would be perfectly legal as long as you claim you might possibly train an LLM with it.

No worries if you don't use this book for training — Meta also pirated thousands of books it never trained with. They downloaded entire libraries from LibGen and torrent sites, had contractors read them (allegedly for "annotation"), then decided not to use many of them. And they never deleted the copies.

Judge Chhabria dismissed key copyright claims, essentially ruling that if the AI outputs don't reproduce the books verbatim, the unauthorized copying and retention is fine.

Why would they pay authors when they can just provide material support to organizations linked to piracy and copyright infringement?

This website is satire demonstrating the absurdity of allowing billion-dollar tech companies to build commercial AI products on stolen content while claiming "transformative use."

If Meta's framework is legally sound, then anyone can download any book from any piracy site, read it completely, keep it forever, and face no consequences — as long as they whisper the magic words: "I might train an AI with this."

The Real Issue: Copyright law is being interpreted to protect tech giants while leaving authors with no recourse. If this stands, it kills the incentive to create original works.