News Corp and Dow Jones file copyright lawsuit against Meta over AI training data

News Corp, owner of The Wall Street Journal and Dow Jones publications, filed a federal copyright infringement lawsuit against Meta Platforms on Tuesday.

The publishers accused Meta of systematically copying millions of articles from their sites to train the Llama artificial intelligence model without permission or payment. The complaint, lodged in New York federal court, seeks damages and an injunction to halt further use of the material.

Executives at News Corp completed the filing after internal audits confirmed extensive scraping of paywalled content dating back to 2020. The suit details how Llama outputs reproduce distinctive phrasing and reporting unique to Journal and Barron's coverage.

Meta implemented defenses claiming fair use protections apply to transformative AI applications, but the publishers rejected the argument as insufficient for commercial models generating revenue through ads and enterprise licensing.

Related cases advanced parallel to this development, though publishers isolated claims to training data ingestion rather than output generation. News Corp appointed outside counsel specializing in IP disputes to lead the effort.

The move followed failed negotiations where Meta offered broad licensing deals below industry benchmarks set by prior OpenAI arrangements. Publishers demanded retroactive compensation scaled to Llama's deployment across Meta's platforms.

Financial impacts emerged immediately, with ad-supported AI queries diverting traffic from original sources. Borrowers in media debt markets adjusted covenants anticipating prolonged litigation timelines.

Resolution timelines extended into 2027 based on docket schedules, as judges consolidated similar publisher actions against AI developers. News Corp confirmed commitment to the suit regardless of settlement overtures.