Meta changes terms to let businesses scrape news sites for AI training, effective Oct. 30
Meta announced that its commercial terms will be updated to permit businesses using Facebook, Instagram, Messenger and WhatsApp to scrape news websites for training artificial‑intelligence systems, with the changes taking effect on October 30, 2026.
The revision matters because it could give the tech giant and its partners broad permission to harvest and store news content—including reporting, photos and other original work—without explicit consent from publishers. Legal analysts say the language may expose news organizations to copyright infringement risks and threaten revenue streams that rely on content ownership.
Under the new language, Meta and third parties acting on its behalf may employ automated tools to access, collect, analyze and retain material from websites linked to business accounts on its platforms. The policy applies uniformly across Meta’s suite of services, extending the reach of data collection beyond traditional advertising or analytics uses.
Experts who reviewed the terms for Nieman Lab warned that the provision could put news outlets in a difficult position, as their material could become available for use in AI models without the specific agreements that normally govern such reuse. The concern is that training data derived from news sites could be incorporated into proprietary AI products, potentially diminishing the value of the original journalism.
Copyright advocates have flagged the update as a possible violation of existing intellectual‑property laws, noting that many jurisdictions require explicit licensing for large‑scale text and image harvesting. The change also raises broader questions about the balance between technological innovation and the rights of content creators.
Regulators in several countries have previously examined Meta’s data practices, and the new terms are likely to draw fresh scrutiny from antitrust and privacy authorities. Industry groups representing publishers have signaled intentions to challenge the amendment, arguing that it undermines the sustainability of the news ecosystem.
As the October 30 deadline approaches, news publishers are evaluating legal strategies and technical safeguards to protect their work. The outcome of any forthcoming legal challenges could shape how major platforms handle AI training data in the future.
