A collective of writers, among them Ta-Nehisi Coates and Sarah Silverman, have charged Meta’s CEO Mark Zuckerberg with endorsing the use of stolen copyrighted content to train the firm’s AI models, as revealed in a recent court document. The lawsuit, disclosed on Wednesday in California, contends that Meta tapped into the LibGen dataset—a Russian-based repository encompassing a large assortment of pirated texts—to develop its Llama AI.

According to legal papers, Meta is said to have acknowledged removing copyright notices from the start and conclusion of academic journal articles, which the lawsuit asserts was designed to hide the company’s engagement with copyrighted works. As reported by *Engadget*, the suit posits that this intentional act was executed to mask Meta’s dependence on pirated materials.

The filing additionally underscores Meta’s worries regarding public image, with *The Guardian* noting that internal communications cautioned: “Media coverage suggesting we have used a dataset we know to be pirated, such as LibGen, may undermine our negotiating position with regulators.” Nevertheless, Meta has persisted in enhancing the integration of its Llama AI within its applications and services.

This legal dispute coincides with other notable transformations at Meta. Recently, Zuckerberg declared the substitution of conventional fact-checkers with Community Notes, a decision that has faced backlash. Furthermore, the company has eased certain limitations on discriminatory and hateful discourse across its platforms and has amplified the promotion of political content on Instagram and Threads.

Meta’s initiatives, including its purported use of pirated materials and policy modifications, are prompting inquiries about the organization’s ethical standards and its broader influence on the digital realm.