Meta removes AI training reference from revised web scraping terms

Meta has removed the word “train” from revised terms governing how it crawls websites linked from business pages, after legal experts warned the policy could allow broad use of that material for AI training. The change comes ahead of the terms taking effect on 30 October, but critics say the revisions leave important questions about the scope of Meta’s scraping rights unanswered.

The earlier terms said Meta could use “any content” to train and support its products; the revised wording allows it to periodically collect website updates for advertising and business products, among other purposes. Meta says the changes do not give it new rights and that site owners can use robots.txt to signal preferences, but the terms do not promise to follow those instructions. Experts argue that broad phrases such as “develop” business products could still cover AI development and potentially extend to entire websites.