Meta, the company behind Facebook and Instagram, used a massive pirate library called “Books3” to train its artificial intelligence models. Inside were copyrighted works published by major publishing houses and independent authors—without any consent, without any compensation, without any respect.
This isn’t a generic accusation. It’s all laid out in black and white in the documents from the ongoing lawsuit against Meta. The Books3 database had already been removed from the internet, but Meta had downloaded it beforehand and used it to fuel its AI.
The books were turned into data. Every paragraph, every idea, every personal style was reduced to numbers and then processed by a system that now writes text on demand—even texts that mimic the very authors who were robbed.
We’re talking about a systematic theft of intellectual labor, the misuse of protected content, and a complete lack of safeguards for those who write, create, and publish.
Many authors have signed an open letter demanding transparency, rules, and justice. But so far, they’re met with a wall of silence and power.
We’re at a historic crossroads. If the future of artificial intelligence is built on stealing the past, then it isn’t innovation—it’s exploitation.
And, as always, the ones who pay the price are the most vulnerable: the authors, the small publishers, and the readers.
If we let AI grow unchecked, without limits or respect, the next book we read might not be written by a person at all, but by an algorithm trained to copy them.
And the fact that all the competitors do the same is no excuse.
