The two settings that quietly remove a website from Copilot and Google's AI answers

What do noarchive, nocache and nosnippet actually do?

This is the rare case where the vendors are explicit. Bing's Webmaster Guidelines state that "NOARCHIVE prevents content from being used in Copilot responses and grounding results," that "NOCACHE limits Copilot to using only the URL, title, and snippet, reducing citation depth and answer quality," and that "DATA-NOSNIPPET and NOSNIPPET prevent Bing from displaying captions and may limit Copilot citation quality." Google's robots meta tag specification says "the noarchive rule is no longer used by Google Search," and its guidance on AI features names nosnippet as a control that prevents a page being used as input for AI Overviews and AI Mode.

Why does nobody notice?

Because the tools everyone looks at measure rankings, and these directives do not touch rankings. A page can sit at position three on Google, show up first on Bing, and still be uncitable by Copilot and excluded from Google's AI answers. Bing says as much about the measurement problem in general: "A decline in clicks does not always indicate a loss of visibility." The reverse is also true. Healthy clicks do not prove the assistants can use you.

There is a second reason. Since March 2026 Google honours robots directives anywhere in the HTML, not only in the head, so a stray tag injected by a plugin, a theme or a script now counts. A directive one developer added to one template in 2019 can be quietly excluding every page built on it today.

How do you check your own site in two minutes?

Open any page, View Source, and search for noarchive, nocache and nosnippet. Then check the response headers, because the same directives can travel in an X-Robots-Tag header that never appears in the page:

curl -sSI https://yourdomain.ca/ | grep -i x-robots-tag

Do this on one page from every template your site uses, not on the homepage alone. The homepage is the page most likely to have been hand-tuned; the product and service pages are the ones built from a template, and the template is where these settings hide.

What if you genuinely need to limit what gets quoted?

Use the tools that do what you mean. Bing offers a positive alternative, the data-snippet attribute, "to specify allowed content" rather than banning quoting outright. And if a page truly must not appear in an assistant at all, the right tool is noindex, which Bing names for exactly that purpose. What does not work is robots.txt, which in Bing's words "controls crawl access, not indexing." A blocked crawler cannot read a noindex it is not allowed to fetch, so the URL gets listed anyway from links elsewhere, with no title and no description.

What we do about it

Finding and removing these is one of the six items in the AI Visibility Pass, and it is checked on every page, not sampled. It sits alongside proving that each crawler can reach the site in the first place, which is the subject of a five-minute check anyone can run, and getting the site into Bing's index, which is why your website needs Bing to appear in ChatGPT. The honest limit stands, in Bing's own words: "SEO does not guarantee rankings or traffic, and GEO does not guarantee grounding or citations in AI experiences." Removing a directive that forbids citation is not a promise of one. It is the difference between eligible and excluded.