Technology & Science
Publishers & Scott Turow Sue Meta, Alleging Zuckerberg Ordered Pirated Books to Train Llama
On 5 May 2026, five major publishers and novelist-lawyer Scott Turow filed a Manhattan class-action claiming Mark Zuckerberg killed licensing talks in April 2023 and instead authorized mass ingestion of pirated texts to build Meta’s Llama AI.
Focusing Facts
- Complaint says Meta torrent-downloaded 267 terabytes of copyrighted books—hundreds of millions of works—to feed Llama.
- Plaintiffs seek statutory damages, a permanent injunction, and destruction of all infringing training copies.
- Anthropic settled a parallel author lawsuit for $1.5 billion in 2025 after mixed early court rulings on AI ‘fair use’.
Context
Copyright law has repeatedly buckled when new technology makes copying effortless. Napster’s 1999–2001 music-sharing frenzy, Google’s 2004 Book-Scan project, and Sony’s 1984 Betamax victory each forced courts to redefine ‘fair use’. The Meta suit underscores a century-long trend: each major information platform—radio in the 1920s, cable TV in the 1970s, search engines in the 2000s, now generative AI—initially grabs content first, argues public benefit later. What is novel here is the scale (hundreds of terabytes) and the allegation that a single executive personally overrode licensing markets, echoing early Standard Oil tactics of vertical integration. On a 100-year horizon, the case may decide whether creative IP remains a licensable commodity or is relegated to raw data for machine learning, reshaping incentives for human authors much as the 1710 Statute of Anne once did for printers and writers.
Perspectives
General-interest newspapers and broadcast outlets
e.g., NPR, The Philadelphia Inquirer, amNewYork — Present the lawsuit as evidence that Meta brazenly stole millions of books, with Zuckerberg allegedly directing one of the largest copyright violations ever, threatening authors’ livelihoods. Stories lean heavily on plaintiffs’ rhetoric of “massive infringement” and “stolen words,” giving scant space to legal precedents that previously favored Meta, which heightens drama and reader outrage.
Tech-industry and business trade press
e.g., CNET, PYMNTS.com, MediaPost — Cast the suit as the latest skirmish in an evolving legal gray zone, stressing earlier court wins for Meta and Anthropic and noting that AI training on copyrighted data can be ruled fair use, so the outcome is uncertain. Coverage foregrounds the ‘transformative innovations’ narrative and prior favorable rulings, reflecting an audience and advertiser base tied to the tech sector and potentially soft-pedaling the scale of alleged infringement.
Right-leaning opinion media
e.g., Breitbart — Frames the case as proof that Big Tech elites like Zuckerberg knowingly trample property rights in their pursuit of AI dominance, echoing plaintiffs’ claim of a colossal theft. Article deploys emotive language and folds the story into broader cultural critiques of Silicon Valley while promoting its own book on the ‘AI war,’ suggesting ideological and commercial motives.
Like what you're reading?