Days ago, The Atlantic dropped an article about how Meta decided to "get books ASAP," in order to train their Large Language Model AIs (https://www.theatlantic.com/technology/archive/2025/03/libgen-meta-openai/682093/). To do that, they went to LibGen to get those books. The problem is that LibGen hosts pirated works.
Disclaimer: I'm not really here to talk about piracy. There are arguments for and against pirating, and I have opinions, but that's not this post.
In the most basic of terms, all Large Language Models are trained by reading and analyzing text and compiling probability metrics that are most likely to be statistically "correct" for any prompt. To keep the models growing, they feed a lot of data into the ever hungry maw of the LLM.
Meta (and we should just assume the others, too) decided to take books for free from LibGen to feed the machine, choosing not to notify or compensate those authors. This shouldn't be a big ask from a company worth $1.46 TRILLION dollars at the time of this post, but they decided to do it under the table. Odds are they're probably not the only ones.
META should be held accountable, but it probably won't be. The other AI companies should be audited to see how pervasive this is, but they won't be.
We couldn't find our books on LibGen, so it looks like we were not personally affected by this. That doesn't make us any less angry at massive tech companies for knowingly stealing from authors just to buff their profit margin for models that are far from perfect. (https://arstechnica.com/ai/2025/03/ai-search-engines-give-incorrect-answers-at-an-alarming-60-rate-study-says/).
What's to do, then?
First off, you can see if you're affected by using the link at the top of the post.
This is a great bluesky thread about the litigation already pending and how as a class action, odds are you're already covered if your works were stolen: https://bsky.app/profile/meredithmooring.bsky.social/post/3lkvmozwyec2t
Here's a link to the author's guild on that existing page (it's older because apparently this had begun filing well before this new evidence was discovered): https://authorsguild.org/news/ag-and-authors-file-class-action-suit-against-openai/
And the final link is to another author's guild page on some more background, but more importantly, some action you can take, like a template to send to Meta and other AI companies if you find your work in the LibGen breach. Some of these options require an Author's Guild membership, but you're not required to join in order to use some of the resources. Other resources are available on becoming a paid member, but you should look at and decide for yourself if it works for you: https://authorsguild.org/news/meta-libgen-ai-training-book-heist-what-authors-need-to-know/
It's easy to be angry, so be angry for a minute. There's a lot going on out there right now and this is just one more thing to throw on the rage fire. You can be mad, or upset, or whatever emotion helps you to process all this. Take some time for that. It can feel overwhelming and exhausting So take some time for yourself. Read a book. Listen to some music. Watch a show.
We often joke to ourselves that the Empire needs more Lions. Honestly, we could all use a little more help from each other right now. So once you're ready to pitch in, look at ways you can support your fellow artists, let your followers know you appreciate them and feed your soul by donating where you can to promote art and community.
- Wade & Tim