Meta to update commercial terms on October 30 requiring business accounts to let it crawl linked website content, including for training

Meta's consolidated Commercial Terms take effect on 30 October. Business accounts must authorise Meta to use automated crawlers and "test credentials" to access, index, cache and collect content from domains linked to their products or shared through Meta services. The terms contain no opt-out. Meta may use that content to "provide, develop, improve, research, train, and maintain" its products. The text does not mention AI by name. Meta says robots.txt is still honoured and its crawling practices are unchanged. It describes the update as a merger of overlapping terms, including the Ad Creative Generative AI Terms and Product Catalog Terms. The "test credentials" wording is new to the commercial terms.

The people most affected are businesses that advertise or sell through Meta, including publishers, and anyone whose site sits behind a login or paywall. Accepting the terms is a condition of using Meta's business products, so they face a choice between agreeing to broad crawling rights and losing access to those tools. Legal and commercial teams should check what content their linked domains expose, and whether test credentials could ever reach gated material. The practical impact is uncertain for now. Meta says nothing changes in how it crawls, and the experts quoted are unsure a court would enforce the terms as a licence for training.

Claims that Meta will bypass paywalls, override robots.txt or dodge "millions" in licensing fees are speculation about what the wording could allow, not evidence of what Meta does or plans. The antitrust "tying" comparison is a legal opinion, not a finding. The coverage also gives no sign that Meta has crawled gated content under these terms, and the terms never say "AI", so the AI-training framing is an inference from the word "train".

1. 01Whether Meta changes or clarifies the wording before or after 30 October, particularly on test credentials, paywalled content and any opt-out for business accounts.
2. 02Evidence from publishers or site operators, such as server logs showing Meta crawlers using credentials or ignoring robots.txt, that would show the terms are being used as feared.
3. 03Any complaint, lawsuit or regulator inquiry (for example under competition or data protection law) that tests whether the terms can be enforced as a training licence, and whether publishers drop Meta business products in response.