I’m Suing Meta For Pirating My Books (and you should, too)

Information Doesn't Want to be Free. It wants to be Properly Remunerated.

Writers around the world are angry with Meta. The social media conglomerate has been at the forefront of a movement to devalue our labor by transforming our creative works into “content” for its platform. It helped dismantle and devalue formidable publishing houses and cratered the news business that many of us depended on for our livelihoods. For two decades we tacitly bought into their world view—one that promised economic freedom through social media engagement—and spent years carefully building up audiences on Instagram, Facebook (and now Threads). The promise was that our public standing would somehow sell more books. Then, when it was too late to turn back, they changed the rules of the game and tweaked the algorithm to hide any posts that link outside the platform. The promise of social media has been empty all along. Instead we’re living in a world where we are little more than digital serfs beholden to the algorithmic whims of techno oligarchs.

Authors were primed to revolt when The Atlantic released a story last week titled “The Unbelievable Scale of AI’s Pirated Book Problem,” which carefully explained how Mark Zuckerberg personally authorized the mass theft of our works to train his newest abomination: an AI program that further devalues our labor. We put our names into a search tool attached to the article and discovered for the first time that our own works were part of Meta’s database. My own name pulled up 22 different pirated works and translations. All told, Meta pirated 267 terabytes of written materials. Or to put it another way, this was the Library of Congress twenty times over.

The mass theft was already being litigated in American courts but the Atlantic story revealed new evidence that was showing up in the discovery process (read the case here). I spent the next few days calling every copyright lawyer I knew, gaming out potential strategies, and seeking opinions from experts in artificial intelligence, and I can see a path where it’s possible to hold Meta (and every other piracy-based LLM) to account. Every affected writer has a chance to recover substantial damages that are commensurate with the unbelievable economic scale of the AI boom. In this letter, I’m going to tell you exactly what I discovered after talking to five different lawyers and how you can pursue your own case against Meta, or maybe even join mine.

But first a little more context.

Those of us in the Author’s Guild had been seeing updates about brewing litigation against various AI companies for more than a year. The Guild warned that AI was a serious threat to the writing profession and was actively seeking ways for authors to get paid to license their works to AI companies. Their efforts began to make some obvious headway in March 2025 when thousands of authors received emails from HarperCollins stating that the Author’s Guild had reached a settlement with an unnamed “AI Company” to use their books as training materials. I was offered $10,000 split between me and Harper Collins to effectively forgive AI for already pirating two of my books. Something didn’t sit right with me about the legalese, so I opted not to sign the contract.

Copyright law in the United States is quite specific on the penalties that come with illegally downloading and distributing creative material. If a work is registered with the copyright office within three months of publication the rights holder is entitled to collect statutory damages of $750 - $30,000 per infringed work. However, in cases of willful infringement where “someone knowingly uses a copyrighted work without permission,” or “acts with reckless disregard for the copyright holder's rights,” it can lead to “increased penalties, potentially up to $150,000 per infringed work.” It seems obvious to me that Meta’s mass downloading of written materials counts as “willful”.

By the most conservative accounting, The Atlantic’s database indicates that I have no fewer than 5 books that meet the criteria for statutory damages. That means Meta owes me between $3750 to $750,000 just for those works. Damages could rise to $3.3 million if a federal jury decides that all 22 of my books, articles and translations that Meta used should be treated the same way. You can check to see if your own books are properly registered by searching the US Copyright Database. On March 21, 2025, Meta celebrated 1 billion downloads of Llama 3. Training the program cost more than $720 million on hardware alone. Even with a penalty of $150,000 per work, it would still only cost them $0.00015 per download. That doesn’t seem unreasonable to me.

Even so, some might balk at such

high payouts for the crime at hand, but there are absolutely precedents. Anyone who was around during the heyday of Napster probably remembers receiving incredibly threatening emails from their internet provider saying that every illegally downloaded song could result in penalties of up to $150,000. While those warnings seemed hyperbolic, the RIAA aggressively pursued litigation against more than 30,000 casual downloaders. Some defendants saw no way out but to end their own lives. Meanwhile people who actually uploaded copyrighted materials for other people to download met with even more serious opposition. In 2012, Reddit founder and internet activist Aaron Swartz downloaded the entire J-STOR academic database from a server at MIT, with the intention of distributing scholarly articles online for free. The FBI caught him and then sought $1 million in penalties and 25 years in jail. Swartz committed suicide before his case went to court.

The Atlantic estimates that Meta used at least 7.5 million books to train Llama 3. If the court awards maximum statutory damages, their total liability comes out to a staggering $1.12 trillion dollars. Meta is currently valued at $1.67 trillion.

If they do pay that full amount, they would still have lost less than Swartz ultimately did. In a just world, it would be possible for Meta to pay out the claims in shares of the company and completely reverse the superstructure between writers and our tech overlords. Which brings me to my next point.

There are many ongoing litigations against Meta and other AI companies for a litany of potential violations. Some tech apologists argue that artificial intelligence doesn’t reproduce copyrighted work, and instead simply ingests the information in a way that is akin to what happens when a human trains at university. My own college reading list included writings by Jon Krakauer, Amitabh Ghosh and Toni Morrison (all of which I paid for at my college bookstore or borrowed from a library, which in turn had to pay for the books). I don’t believe I owe those authors payment for whatever inspiration they contributed to my own writings. The thorny question of the origins of creativity is still very much alive in philosophical circles. It would be a tall order for a court to resolve it during litigation. Luckily, none of the ongoing cases require resolving it.

This is because the lawsuits against Meta, OpenAI and other AI companies don’t rest on philosophical arguments, but on the basic facts of how the companies acquired the pirated material in the first place. Just like the cases against Napster users filling up their music collections for free in 1999, Meta’s violation occurred when they decided to willfully and knowingly use pirated materials to train their LLM. According to discovery materials acquired in the first lawsuits against Meta, engineers involved in training the LLM, their phalanx of lawyers, and even Mark Zuckerberg signed off on acquiring training material from illegal pirate databases. Downloading from those websites was a crime all on its own.

Meta knew that their LLM was not as good as products by other companies and specifically sought out books that could provide higher quality training data than simply scraping the web and/or user-submitted writing on their own platform. They spent several months negotiating AI-training rights with HarperCollins and Penguin-RandomHouse. They even signed deals with a handful of African publishers. But ultimately they decided that it would be cheaper and faster to simply download books illegally. One Meta employee wrote that pirating was actually better because, “The problem is that people don’t realize that if we license one single book, we won’t be able to lean into fair use strategy.” A research team at Berkeley statistically proved that training on copyrighted material improved Llama’s performance by more than 7% (lawyers take note: the same research team also outlined a methodology to determine if a specific book was used in the training set).

Meta spent 2 months downloading hundreds of terabytes of pirated books from torrenting websites LibGen and Anna’s Archive. They accessed a database called Books3—one that other AI companies had also used—that contained more than 175,000 copyrighted books. Importantly, while torrenting the massive archive of literary works, Meta used a bittorrenting protocol that not only illegally downloaded materials (what torrenters term “leeching”) but actively uploaded it for other people to access as well (“seeding”). This means that even before Meta loaded a single book into their LLM training scheme, they had already both stolen and distributed your copyrighted material.

Meta seems to believe that while downloading a single book might well be a crime, downloading millions of them to build a for profit business model of their own is somehow not. The logic is akin to arguing that while murder is wrong, genocide is just good business.

You can read all of the details of the case against Meta in the motion for a partial summary judgement filed on behalf of a small group of writers. It’s riveting reading for any writer who wants to feel righteously incensed at the scope of Meta’s theft.

But now I want to turn to the real topic of this piece and tell you, my fellow writer, what you can do to take this case right to Meta’s doorstep. Because unlike decades past we finally have the power to do something about the incredible greed of the techno-oligarchy. They have violated the exact same laws that they used to prosecute small time music fans and internet activists.

So how can you sue Meta?

The good news is that if you do absolutely nothing, you will still likely get the opportunity to join a class action lawsuit against Meta in the next few years without doing anything at all. Matthew Bitterick’s suite of lawsuits against various tech companies are currently wending its way through the San Francisco courts. Though he initially filed on behalf of Richard Kadrey, author of the Sandman Slim series, and the comedian Sarah Silverman, he suspects that the court will soon combine his and other similar legal actions into what might be the largest class actions lawsuit of all time (the previous record holder occurred in 1998 against Big Tobacco was worth just over $200 billion). Once it becomes a class action, all eligible writers can simply apply to be part of the case. If you subscribe to this newsletter, I’ll let you know when that happens.

While this path is open to just about every writer, there are some notable disadvantages to it. By participating in a class action, your individual claims will likely become diluted down and shared across the entire spectrum of litigants. The lawyers may seek a settlement that you ultimately disagree with. From Meta’s and the court’s perspective, a class action is much easier to deal with than thousands of smaller cases tried in front of an endless assortment of juries. However, you always have the option to opt-out of a class action and pursue the case on your own.

In the coming weeks I plan to do just that.

I recently retained a lawyer to represent me along with a small group of similarly situated authors. Our goal is to extract the maximum possible damages from AI companies that have illegally used our works. Since copyright violations are entitled to a jury trial, we will seek to plead our case in front of twelve honest citizens. It would be amazing (if unlikely) to depose Mark Zuckerberg himself. Furthermore, we will also pursue litigation against every other AI company that used Book3, LibGen, Anna’s Archive, or any other pirate database to train their artificial intelligence programs on author works. We will pursue separate jury trials in each instance..

This will likely be a long road of litigation, but also a necessary one to protect our work from technological strip mining. We will likely file cases against OpenAI, Google, and Elon Musk’s Grok.

Of course, you are probably wondering where a small group of authors was able to gather the legal war chest necessary to take on the largest technology companies in the world. The fact of the matter is that the potential damages are so high that many lawyers have already expressed interest in representing us on contingency. This means that our legal team will collect 30% of the winnings. And, since copyright cases often allow the plaintiffs to recover legal fees in the judgement if we prevail, there is a good chance that those very technology companies will end up footing the bill for us. Much like the asbestos litigation over the last few decades, we expect that our case can be a blueprint for other writers to follow.

Now maybe you want to know how you can get involved in your own litigation or perhaps join in my effort. Many people have already reached out to me directly, and I welcome correspondence with other writers who have been wronged by Big Tech. I might be able to fit one or two more people into my own suite of lawsuits. I would love to be able to play a role in connecting lawyers interested in representing writers on contingency with writers who desperately want to protect their copyrights.

However, there is also no reason that you can’t organize your own legal action. Do you remember that friend of yours from college who went on to a storied legal career? Call them up and see if they have advice for who might represent your claim. Connect with other writers who are equally furious. And, most importantly, pick up your pen and start doing what you do best: write. If we make this loud, we will gain the benefit of public sentiment. We will bring the tech oligarchs to court and convince juries of our peers that the technologists were wrong: information doesn’t want to be free. It wants to be properly remunerated.

I will be keeping you abreast of my fight against the technocracy here on Substack and on my YouTube channel over the next few years. I cover a wide variety of topics—from political rabble rousing, to longform investigations of health grifters and oligarchs. If you want to follow along, please consider subscribing for free. I would also sincerely appreciate anyone who decides to sign up for a premium membership here on substack (it’s just $8/month). It would really mean the world to me.

©PokeyBear LLC

Below this paywall is an embedded video that is already free on YouTube titled “How Writers Can Sue Zuckerberg into Oblivion.” I'm including it to make it easier for you to become a paid subscriber as upgrading can be tricky on Substack.