Meta’s new business terms, now delayed to Nov 20: what they mean for your website

Meta’s updated Commercial Terms now take effect on November 20, 2026. On October 23, Meta pushed them back from October 30 after publishers and advertisers complained to regulators, but it hasn’t withdrawn them, and the clause that worries site owners is still in the text. It lets Meta crawl, index and collect content from websites “associated with” any business that uses Facebook, Instagram or WhatsApp business tools, and you accept it simply by keeping those tools. Meta already removed the explicit reference to training on that content and says the crawling applies only to its ads and business products. If you have a Facebook Page or run Meta ads and also own a website, use the extra three weeks to decide which Meta crawlers your robots.txt allows.

Below: what the terms say, how they changed since October 15, who counts as a business account, what each of Meta’s crawlers does, and how to set robots.txt on a WordPress site.

What Meta’s new commercial terms say

The updated terms replace the version dated March 6, 2024. Nieman Lab reported on October 15 that the new text authorises Meta, or anyone acting for it, to use automated software (“web scrapers, bots, robots, spiders, crawlers and user agents”) to access, index, cache, analyse, collect, retrieve or crawl domains and web pages “associated with your products or services”, and to periodically pull updates from them.

The version Nieman Lab reviewed also said Meta could use that content to “provide, develop, improve, research, train” its products, with “AI at Meta” among them, and allowed Meta to create test credentials to access covered sites. PPC Land’s clause-by-clause review notes other changes too: new rules for AI features in ads, a broader warranty disclaimer, and a clause letting Meta discontinue products at its discretion.

There is no opt-out. The terms count continued use of Meta’s business tools after the update notice as acceptance. The alternative they offer is to stop using Meta products for business purposes.

What changed after the backlash

• October 15: Nieman Lab publishes the clause with comments from three legal experts, who say it could cover training Meta’s AI on publishers’ content.
• October 16–17: Meta revises the text and removes the word “train” from the permitted uses (Quartz, Android Headlines).
• Meta’s position: “the crawling terms we consolidated only applied to our ads and business products,” and the change “does not change how Meta uses web crawlers” (Social Samosa, citing The Verge). A spokesperson told Nieman Lab that site owners can still manage access with robots.txt.
• Critics’ position: the News/Media Alliance says the terms still give Meta broad permission to crawl external sites, and legal experts quoted by Nieman Lab argue a signed contract could be used to override robots.txt instructions.
• October 23: Meta moves the effective date of the Commercial Terms and Self-Serve Ad Terms to November 20, 2026, saying it wants more time to explain the changes. It’s a delay, not a withdrawal: the crawling clause is unchanged. The European Commission welcomed the delay (PPC Land, Nieman Lab).

Two things remain unclear: whether Meta changes the wording again before November 20, and how it defines a domain “associated with” a business. A Page that links to your website is the obvious case.

Does this apply to you?

Nieman Lab’s definition of a business account is broad: managing a Facebook Page or group, running ads, sending marketing messages, or using Meta’s business dashboard. A freelancer with a Facebook Page that links to a portfolio site fits that description. So does a shop that sends order updates through WhatsApp Business.

If your site holds anything you wouldn’t want copied into another company’s systems, such as paid guides, client-only pages or draft pricing, this is a good moment to move it behind a login. Robots.txt is a request, not a lock. For the same reason, it’s worth checking what your AI tools do with your files; our guide on keeping client data out of AI training covers ChatGPT, Claude, Gemini and Copilot.

Meta’s web crawlers and what blocking each one does

Meta lists five crawlers on its web crawlers page for webmasters. They do different jobs, so blocking all of them has side effects.

User agent	What Meta says it does	Follows robots.txt?	If you block it
facebookexternalhit	Fetches pages shared on Facebook, Instagram or Messenger to build link previews	Not always (may bypass for security checks)	Shared links lose their title, description and image
meta-externalagent	Crawls to train foundation AI models or improve products by indexing content	Yes	Opts you out of AI training crawls; also out of this indexing
meta-webindexer	Builds the index Meta AI uses to answer and cite search queries	Yes	Meta AI is less likely to cite or link your pages
meta-externalads	Crawls to improve ads and other business products	Yes	Effect on your own ads isn’t documented
meta-externalfetcher	Fetches a specific link when a user asks Meta AI to	Not always (user-requested fetches)	May still fetch pages a user asks about

For most small businesses, a sensible middle ground is to allow facebookexternalhit (so your links still look right when shared) and decide on meta-externalagent based on how you feel about AI training. Note that Meta uses that single token for both training and indexing, so you can’t accept one and refuse the other.

How to block Meta’s AI crawler with robots.txt

Add these lines to the robots.txt file at the root of your site (yoursite.com/robots.txt). This example blocks the AI training crawler and the ads crawler and leaves link previews working:

User-agent: meta-externalagent
Disallow: /

User-agent: meta-externalads
Disallow: /

On a WordPress site with Yoast SEO, you can edit the file from Yoast SEO > Tools > File editor if your host allows file editing; otherwise edit robots.txt through your host’s file manager. Then:

1. Open yoursite.com/robots.txt in a private browser window and confirm the new lines appear.
2. Allow up to 24 hours for the change to take effect. Meta’s documentation says its crawlers can cache robots.txt.
3. Share one of your pages in a private Facebook post or Messenger chat to confirm link previews still load.

If your site moved recently and you’re already editing DNS or host settings, check your email records at the same time; our guide to SPF, DKIM and DMARC explains what to look for.

FAQ

When do Meta’s new commercial terms take effect?

November 20, 2026. They were due on October 30, but Meta announced a delay on October 23. Using Meta’s business tools after November 20 counts as accepting them.

Can I opt out of Meta’s new commercial terms?

Not partially. Continuing to use Meta’s business tools after November 20, 2026 counts as accepting the terms. The only way to refuse is to stop using Meta products for business purposes. You can still limit which Meta crawlers your site allows through robots.txt.

Did Meta drop AI training from the terms?

Meta removed the word “train” from the clause about content collected from associated websites around October 16–17, 2026, and says the crawling only applies to its ads and business products. Critics say the remaining language is still broad. Meta-ExternalAgent, which Meta documents as crawling to train AI models, still exists and is controlled through robots.txt.

Will blocking Meta crawlers break my Facebook link previews?

Not if you leave facebookexternalhit allowed. That crawler builds link previews; the AI and ads crawlers use different user agents.

Does this affect my personal Facebook or Instagram account?

The commercial terms apply to business use: Pages, ads, marketing messages and business tools. A purely personal account isn’t covered by them, though Meta’s general terms and privacy policy still apply.

How we checked: we read Meta’s preview of the updated terms and compared it with Nieman Lab’s original report (October 15, 2026), PPC Land’s clause review, reporting on Meta’s revision from Quartz, Android Headlines and Social Samosa, coverage of the October 23 delay, and Meta’s developer documentation on its crawlers. Last checked: October 27, 2026; date references updated for the November 20 delay. This is general information, not legal advice.

Compare free social media schedulers, including Meta’s own.

Related guides

• PHP Security Release Oct 21: What WordPress Owners Should Do
• How to Check If a Website Is Down for Everyone
• Google Privacy Policy Change: Keep Gmail Out of Your Ads

Sources

• Meta Commercial Terms (preview of the update)
• Meta for Developers: Meta web crawlers
• Nieman Lab: Meta’s new commercial terms will allow the scraping of news sites (Oct 15, 2026)
• PPC Land: Meta’s new terms let it crawl business websites from Oct 30
• Quartz: Meta removed “train” from its business terms (Oct 19, 2026)
• Social Samosa: Meta removes terms that raised scraping and AI training concerns
• Android Headlines: Meta backtracks on AI scraping clause (Oct 17, 2026)
• PPC Land: Meta delays website crawling terms to Nov 20 after publisher complaints (Oct 2026)
• Nieman Lab: Meta delays rollout of terms that would allow broad scraping of news sites (Oct 2026)

Next step: open yoursite.com/robots.txt now and check whether it mentions any Meta user agent. If it doesn’t, decide on meta-externalagent before November 20.