Menlo Park — Meta's New Business Terms May Allow AI Training on Sites

Meta, the Menlo Park, California-based parent company of Facebook and Instagram, is updating the commercial terms of service for business accounts on Oct. 30, and the new language gives Meta broad permission to scrape and train artificial intelligence on content linked through those accounts, according to Ars Technica, which cited reporting from Nieman Lab published Oct. 16, 2026. Businesses cannot opt out. The terms instead suggest that companies stop using Meta services if they disagree.

The change matters most for news organizations, which routinely link articles from Facebook and Instagram business pages. If enforced broadly, the terms could let Meta avoid paying licensing fees that publishers currently seek to charge AI companies for using their journalism to train chatbots and other models.

What Is Meta Changing In Its Commercial Terms?

The updated terms authorize Meta, or any company acting on its behalf, to use automated software such as web scrapers, bots, spiders, and crawlers to access, index, cache, collect, and periodically pull updates from any domain, web page, or URL tied to a business account, according to the terms language quoted by Ars Technica. The text goes on to say Meta can use that content to "provide, develop, improve, research, train, and maintain the integrity of" its products and services.

The word "AI" does not appear anywhere in the clause. Legal experts told Nieman Lab the AI connection comes from a single use of the word "train" combined with the scraping and crawling language.

Who Does The New TOS Affect?

Every business account on Meta's platforms falls under the new commercial terms, not just news outlets. But Nieman Lab said the update is especially significant for journalism because nearly every news organization, including Nieman Lab itself, maintains at least one Meta business account.

"While journalism isn't the only industry that will be affected by this update, virtually every news organization has at least one Meta business account, including Nieman Lab," the lab wrote, as quoted by Ars Technica.

News publishers have spent the past several years trying to negotiate licensing deals with AI companies or suing over unlicensed training. OpenAI is currently fighting publishers in court over whether training AI models on news articles without payment counts as fair use. Meta's new commercial terms could let the company sidestep that fight entirely by getting consent baked into account sign-up, regardless of how courts eventually rule on fair use.

Does This Override Robots.txt Blocks?

Robots.txt is a standard file website owners use to tell web crawlers which pages they can and cannot access. News sites commonly use it to block AI training bots.

A Meta spokesperson told Nieman Lab the company has no plans to use the new commercial terms as legal grounds to ignore robots.txt instructions blocking scraping on news sites. A Meta spokesperson also told Ars Technica directly that "there is no change to how Meta uses web crawlers; any suggestion to the contrary is false." Meta says it still allows news organizations to block its crawlers and provides instructions for doing so.

That said, legal experts who spoke with Nieman Lab said the question is what happens if a court eventually rules that AI training on non-news content is fair use. In that scenario, publishers worry Meta's position could shift, and the new terms would already be in place to argue that robots.txt blocks do not matter because businesses agreed to scraping when they set up a Meta account.

Why Are Legal Experts Worried?

Three legal experts interviewed by Nieman Lab said it remains unclear whether a court would actually enforce these terms against a publisher that tried to block scraping elsewhere. But those same experts said the language could be used to argue that Meta has the right to scrape news sites and use the content for AI training, according to Ars Technica's reporting.

What Happens Monday, Oct. 30?

The new commercial terms take effect Oct. 30, 2026, with no opt-out mechanism built into the sign-up or renewal process for business accounts. Companies that do not want to accept the terms have one option under the current language: stop using Meta's commercial services.

For now, Meta says its crawling practices are not changing and that robots.txt files will still be honored for news publishers. Whether that holds if a fair-use ruling favors AI companies is the open question lawyers, publishers, and outlets including Nieman Lab say they are watching closely.

By The Numbers

• Oct. 16, 2026 — Date Nieman Lab first reported the TOS change, as cited by Ars Technica.
• Oct. 30, 2026 — Date Meta's updated commercial terms take effect for all business accounts.

For related coverage of Meta's AI push, see HTT News' earlier report on Meta's new AI-focused subscription plans.