Meta Imposes Oct. 30 AI-Training Terms on Business Sites, Blocking Opt-Out
Updated
Updated · Ars Technica · Oct 16
Meta Imposes Oct. 30 AI-Training Terms on Business Sites, Blocking Opt-Out
3 articles · Updated · Ars Technica · Oct 16
Summary
October 30 is the trigger date for Meta’s commercial-terms update, which would require business accounts to let Meta crawl and use linked website content to “train” and improve its products.
The revised terms offer no opt-out and instead tell businesses that reject the policy to stop using Meta services, according to legal experts and Nieman Lab.
News publishers face the sharpest impact because the language could give Meta broad access to articles that might otherwise command millions in AI licensing fees.
The wording does not explicitly mention AI, but it authorizes scraping, crawling, indexing and periodic updates from associated domains, and includes content use for research and training.
The move lands as news organizations are already fighting OpenAI in court over whether training on copyrighted journalism without licenses qualifies as fair use.