Close
Current temperature in Boston - 62 °
BECOME A MEMBER
Get access to a personalized news feed, our newsletter and exclusive discounts on everything from shows to local restaurants, All for free.
Already a member? Sign in.
The Bay State Banner
BACK TO TOP
The Bay State Banner
POST AN AD SIGN IN

Trending Articles

Is AI a key to Black economic mobility or will it increase disparities?

Who’s running for Boston City Council?

New rent control legislation takes aim at drastic increases

READ PRINT EDITION

AI stole my words — then accused me of plagiarism

ReShonda Tate

When I began researching the downside of AI, I decided to run a few chapters from my book “Miss Pearly’s Girls” through one of those popular AI detectors—you know, the kind that schools, editors and employers are using to “verify” whether work is human or machine-made.

The result?

“This work is 92 percent AI-generated.” That’s what it told me.

There was just one problem. I wrote “Miss Pearly’s Girls” in 2020. From scratch. By hand. Long before ChatGPT even existed.

What began as a curiosity quickly turned into a gut punch. How could my own words—written from my spirit, my experience and my culture — be misclassified as artificial? I soon found out the answer is more complicated —and far more disturbing — than a faulty detector.

My words were stolen

After digging deeper, I discovered that my book had been scraped — yes, stolen — and used to help train artificial intelligence models without my knowledge or permission. It turns out that not just that book but 49 of my other books were also used for training. And I’m not alone. Apparently, Meta used the work of thousands of authors to train its AI. That’s right. It fed our work into the AI machine, repackaged and returned to the world without our names attached.

Much of this happened through digital piracy. If you haven’t heard of LibGen (Library Genesis) or Anna’s Archive, let me introduce you to the Internet’s biggest warehouse of stolen books. These platforms are essentially black-market libraries stocked with pirated copies of thousands — maybe millions — of books, academic texts and creative works that authors and publishers never gave consent to distribute.

Most of us writers knew about it. We sighed. We tolerated. We gritted our teeth and kept creating.

But now there’s something new—and far more worrisome.

Meta’s alleged raid on pirate sites

Recently filed court documents in the Northern District of California allege that Meta, the tech giant behind Facebook and Instagram, deliberately targeted LibGen and Anna’s Archive to extract content and train its latest AI model, LLaMA 3.

That’s right. The allegation is that Meta authorized a data raid on a known pirate site. And that raid may have included your work. It definitely included mine.

Legal action is now underway. A class-action lawsuit — Kadrey v. Meta — has been filed, accusing the company of copyright infringement on a massive scale. If your book was part of that training set, you are automatically included in the lawsuit. You don’t need to do anything just yet. The court is first deciding whether Meta broke copyright law. A ruling is expected this summer.

But this isn’t just about Meta.

OpenAI, Microsoft, Anthropic and others are facing similar lawsuits. In the case against OpenAI (another company I discovered was using my books to train AI), the Authors Guild, along with major names like John Grisham, George R. R. Martin and Jodi Picoult, is fighting for all U.S. authors whose work has been ingested into GPT’s training data — whether you’re famous or not.

The impact hits harder for Black writers

Let’s be honest. Black authors already face enough challenges: fewer publishing opportunities, smaller advances and limited marketing support. Now, to add insult to injury, our voices are being absorbed into tools that may ultimately replace or erase us.

And to make matters worse, we’re being accused of using AI when we haven’t.

I’m hearing more and more stories from Black writers being flagged — falsely — by AI detectors. These tools claim to detect “robotic” patterns, but often what they’re really detecting are the nuanced, unique ways we speak, write and structure language.

It’s not just technical bias—it’s cultural erasure.

What you can do right now

If you’re a writer, creative or educator, here are some things you should know and do:

Check if your book is listed on LibGen or Anna’s Archive. Many are, even if you never authorized it.

Read up on the class-action lawsuits. You don’t have to be a named plaintiff to benefit if the cases succeed.

Be cautious with AI detection tools. They are notoriously unreliable and biased.

Use trusted tools like the Authors Guild AI Resource Page to protect your work.

Speak up. The more we raise our voices, the harder it is for tech companies to ignore the real harm they’re causing.

ReShonda Tate is a bestselling and NAACP Image Award-winning author of more than 50 books, some of which have been made into TV movies. She is a former TV news anchor and reporter who has worked in Houston, Oklahoma City, and Beaumont. Tate also works as a ghostwriter, literary editor and consultant, and screenwriter. 

Leave a Reply