Meta's new business terms let it train AI on websites linked to business accounts

Meta is updating its commercial terms in a way that appears to let it crawl and scrape websites associated with the businesses that use its platforms, including online publishers, and use that content to develop, research and train its AI products. The new terms take effect on October 30.

Legal experts have raised alarms because the update does not offer an opt-out. A business that does not want its linked website used for training is effectively told to stop using Meta's services. Critics say the change could let Meta avoid paying the licensing fees that other AI companies have started paying publishers for training data.

Meta says it is simply consolidating its commercial terms. The practical effect for any company with a Facebook or Instagram business account is the same: its website content becomes part of the deal.

The move also lands as publishers and AI companies keep fighting over who has to pay for training data, with several licensing deals signed this year and lawsuits still working through the courts.

Our take

Training data licensing was turning into a real market. Terms of service that make consent a condition of using an ad platform undercut that market without a single negotiation.