Starting October 30, Meta can crawl websites linked to business accounts on Facebook, Instagram and WhatsApp and use what it collects to train its AI products. For newsrooms and small businesses, that means Meta scraping news sites linked to their accounts, with no opt-out short of leaving the platforms.
The change surfaced in legal-expert analysis from Nieman Lab, which published on October 15. Meta had alerted users in an early-October blog post and in Meta Business Suite. Coverage spread quickly across outlets from AI Weekly to NewsBytes, and the timing matters. The effective date is just over two weeks away.
This story relies on the reporting and legal analysis cited here rather than a direct reading of Meta's terms document.
What Meta's updated terms say about scraping news sites
The terms let Meta use web scrapers to crawl content from websites "associated with" business accounts, according to AI Weekly and Nieman Lab. The covered accounts include those that manage pages, run ads, send marketing messages or use the business dashboard. News publishers are included.
The most consequential word is "train." The terms say content collected from those sites can be used to train Meta products, and they define "AI at Meta" as one of those products. Three legal experts who reviewed the terms for Nieman Lab said the language could be read to let Meta scrape news sites and train AI on their content, according to the reporting from NewsBytes.
Media and intellectual property attorney Matt Topic of Loevy & Loevy was blunt. "This is a remarkable provision to bury in [the] terms of service," he said. He also said that "It's not clear whether a court would enforce this, but publishers should assume that it would be enforceable."
Why there is no opt-out before October 30
Business accounts have no opt-out window before October 30, according to the reporting from daily.dev, NewsBytes and Archynewsy. The terms also treat continued use as agreement. A business that keeps using Meta's tools after that date is considered to have consented, and anyone who rejects the terms must stop business use of the products.
| Item | What the reporting says |
|---|---|
| Effective date | October 30, 2026 [1][2][3] |
| Scope | Business accounts on Facebook, Instagram and WhatsApp [1][2] |
| Opt-out before effective date | None [2][3][6] |
| Consent after effective date | Continued business use treated as consent [2][3][6] |
| Notice to users | Early-October blog post and Meta Business Suite [1] |
Meta's own account is murky. A spokesperson told Archynewsy that the crawling language had been scattered across different terms, including those on generative AI in ads, and that the update consolidates it in one place. A writer for Steampunk Explorer said they could not find the new paragraph in Meta's current terms, and that a Meta spokesperson had not produced the earlier language they requested. USA 24 has not been able to confirm either account.
Can robots.txt and paywalls stop Meta's crawlers?
The standard tool for telling crawlers to stay out is robots.txt, a plain-text file at the root of a website. Meta has published robots.txt blocking instructions for its crawlers, but experts question whether they will hold, according to NewsBytes.
The bigger concern is the contract language itself. Legal experts warn the terms could be used to argue that they override robots.txt blocking instructions. That argument is legally untested, and it is unclear whether the clause would also override paywalls, according to daily.dev and NewsBytes.
Who is most exposed: publishers, small businesses and creators
News publishers face the most direct exposure. Nieman Lab reports that virtually every news organization has at least one Meta business account, including Nieman Lab itself. For those newsrooms, staying on the platform after October 30 may be treated as consent to crawling and AI training.
The exposure reaches well beyond media. Any US website owner whose domain is linked to a Facebook, Instagram or WhatsApp business account could be affected, including small businesses and creators. A Steampunk Explorer report aimed at creators raised the same concerns.
The money at stake could also be large. Topic said the content at issue could carry licensing value in the many millions of dollars, according to Nieman Lab and AI Weekly.
How the copyright fight over AI training is shaping up
The Meta terms arrive during a busy stretch for AI copyright cases. On October 8, USA Today Co. and 13 affiliated papers sued OpenAI in Manhattan federal court, alleging that OpenAI copied hundreds of thousands of articles from 19 publications. The suit seeks more than $250 million, according to Law Commentary and Trademark Lawyer Magazine. The complaint also asks a court to order the destruction of models and training datasets built on the plaintiffs' work.
Earlier rulings have been mixed. A court found that training Anthropic's models on copyrighted books was fair use, but storing pirated copies was not. That case then settled for US$1.5 billion, roughly US$3,000 per work, according to Norton Rose Fulbright. In the Meta-specific Kadrey case, the court granted a partial motion to dismiss on fair use for the training claims, while torrenting-related claims remain active.
Meta has also struck paid AI data deals with news publishers. Reuters, via MarketScreener reported deals that included USA Today, but the snippet reviewed did not show the year, so that timing remains unverified.
Who bears responsibility when automated systems overstep is also a live policy question. USA 24 has covered that liability debate in Who Pays When AI Agents Go Rogue?
What to do before October 30
Publishers and business owners have a short window to decide. The steps below are practical starting points, not legal advice.
- Check whether your business account is tied to websites you own, including through linked pages, ads or domains.
- Decide before October 30 whether to keep using Meta's business tools, since continued use may be treated as consent.
- Document your current robots.txt and terms-of-service settings, with dates, so you have a record if a dispute arises.
- Apply Meta's robots.txt blocking instructions, knowing that experts question whether they will hold.
- Ask your counsel whether a written carve-out from Meta is possible.
What to watch before and after the deadline
Three open questions will determine how much this change matters. The first is whether Meta narrows, clarifies or delays the clause. No confirmed Meta response to the Nieman Lab analysis appeared in the sources reviewed.
The second is whether publishers or lawmakers sue or seek a carve-out before October 30. That is not yet known. The third is how Meta's crawlers behave after the date, and whether robots.txt blocks hold up when a site tests them.
The first real test may come from a publisher that finds its content crawled after October 30 and decides to press the point in court. Whatever happens, the outcome will signal how AI companies write training rights into the business tools that small operators already rely on.
