So Meta and OpenAI pirated your books and articles. Here’s what you can do.

The Atlantic published a new search engine that lets writers discover which of their works have been pirated and used to train AI models.

Alex Reisner published another outrage-sparker in The Atlantic this morning.

Reisner is the writer who put together the world’s first publicly searchable database of Books3, the pirated book dataset that Meta used to train its AI model, back in Sept. 2023.

This morning he published a blockbuster followup: A search engine for LibGen, the massive trove of pirated books and articles that Meta and OpenAI use to train their AI models. (The work Reisner has done over the past 18 months has been absolutely critical in exposing the extent of corporate tech’s property theft. Check out his articles here.)

To be clear: Meta and OpenAI haven’t just innocently stumbled into a briar patch of complex legality. Emails that have come to light as a result of copyright lawsuits show that Meta employees knew they were using stolen intellectual property, and some had ethical qualms about it. Top Meta executives told them to press on, damn the legalities.

Reisner’s been following those copyright cases closely, and last night got his hands on the latest docs:

Court documents released last night show that the senior manager felt it was “really important for [Meta] to get books ASAP,” as “books are actually more important than web data.” Meta employees turned their attention to Library Genesis, or LibGen, one of the largest of the pirated libraries that circulate online. It currently contains more than 7.5 million books and 81 million research papers. Eventually, the team at Meta got permission from “MZ”—an apparent reference to Meta CEO Mark Zuckerberg—to download and use the data set.

Reisner and The Atlantic have helpfully posted a search engine to find out if your work is in LibGen.

find your pirated work in this search engine

It’s available here, and you will need to subscribe to The Atlantic to access it. Writing is labor, and we need to pay people for their work.

You’re a theft victim, and you’re angry. what can you do?

Welcome to the club. Here are your options.

You could sue, in theory. Corporate tech knows you don’t have the money to sue them for copyright infringement. That’s what they’re counting on. The writers who have banded together to file federal lawsuits, some with the backing of groups like The Authors Guild, are courageous individuals. Some of them, like Sarah Silverman and Michael Chabon, are successful authors of means. Most writers don’t have the means.

Congress is no help to you right now. Corporate tech has captured the Trump Administration and is lobbying the President to give Meta, Open AI, et al, carte blanche to steal all the data they want.

The only political bodies pushing back against all this, as of 2025, are state legislatures.

State lawmakers from both parties are skeptical of corporate tech and the coming of AI. They recently enacted laws like the Colorado Artificial Intelligence Act, the California AI Transparency Act, and the California Training Data Transparency Act. They have real power, and they listen to people like you. But only if you contact them.

1. Support California’s AB 412 right now. like today.

You’re a writer. Write.

The California Assembly is now considering AB 412, the AI Copyright Transparency Act. This bill addresses the AI piracy problem head-on. States can’t regulate copyright directly (it’s a federal law) but they can enact laws, like this one, that require AI developers like Meta to disclose their use of copyrighted material to train an AI model. Required disclosure is a major step towards ending the piracy crisis.

The tech lobby is actively working to kill the bill. Because AI companies don’t want to pay you for your work. They want to keep stealing it. That plan has worked well for them so far.

Write a letter in favor of the bill. Include your personal experience. Here’s how:
1. Go to this page and register. (It’s essentially a bot filter.) It’ll take you 20 seconds.
2. Once you’re signed in, go here to submit a letter for AB 412. It works best if you write a letter first and save it (Word or .pdf), then upload it via the portal.

Writing a support letter like this really works. We cannot emphasize this enough. It will go in the bill’s file. Committee members read the file.

2. Find a bill in your state and advocate for it.

At the Transparency Coalition we’re working with state legislators on 11 bills in six states this month. There’s probably an AI-related bill in your state that could use your support.

Find it. Write to your direct state representatives. You’re their constituent. They listen to you. We’ve got a bill finder map that will help you discover what’s happening in your state:

Did that? Great. Now email your friends and ask them to do the same. Emails are better than posting on social media, but do that too.

3. REgister with an agency that licenses training data.

This may seem counterintuitive. One of the best things you can do to protect your work is to sign up with an AI licensing platform like Created By Humans.

You don’t have to allow AI developers to use your work to train their models. In fact, you can register your work and specifically say you don’t want it used to train AI.

What matters is that you sign up. Agencies like Created By Humans are protecting everybody’s copyright and intellectual property rights by proving that books have real value as training data. This bolsters the case for those authors currently defending their copyright against AI in court. We’ve got an article about it:

4. support us.

At the Transparency Coalition, we’re fighting for AI safeguards every day. We’re fighting to protect your intellectual property from rampant theft.

We need your financial help to keep doing it. Donate.

We’re not going to Facebook-post our way to progress. At TCAI we have a bias for action. We show up. We testify at bill hearings. We knock on doors and talk one-on-one with state legislators and their staff members. We go through bills line-by-line and mark up the bullshit the tech lobby tries to insert. We turn tech jargon into plain language so everyone can understand how AI works and why commonsense transparency requirements are both possible and necessary.

We do the work. But we can’t continue to do it without your support.

Oh, and subscribe to AI Spotlight, our monthly email newsletter. It’s free. If you read all the way to the end of this article, you’re our kind of people. You’ll like the newsletter. Sign up.