Google today updated "Control what you share with Google Search," the Search Central documentation page that sets out how site owners can keep content out of Google's results. The page now lists seven methods in its main table, adding a row for the robots meta tag, data-nosnippet and X-Robots-Tag, and a row for the "Search generative AI control" in Search Console. According to the Search Central changelog entry dated October 21, 2026, the options were already available and the update reflects "the current set of options available to site owners." The page itself carries a "Last updated 2026-10-21 UTC" stamp.

In Short

Google has a help page that explains the ways a website can stop its pages from showing up in Google Search, and today it added two methods that already existed but were missing from the list. One lets sites control how their text is shown in results, and the other is a Search Console switch that keeps a whole site out of AI Overviews and AI Mode. Nothing new was switched on, but the page is now a more complete map of what each control covers and, by omission, what it does not.

What the changelog says

The entry sits at the top of Google's "Latest documentation updates" page, under October 2026. Its headline reads "Updated the control what you share with Google Search documentation." The explanation is two sentences long.

"Updated the control what you share with Google Search documentation to include existing options such as the robots meta tag, data-nosnippet, X-Robots-Tag, and the Search generative AI control, along with minor clarifications that the page applies to Google Search," according to the "What" line. The "Why" line is shorter still: "To keep the page fresh and reflect the current set of options available to site owners."

Two words in that entry carry most of the weight. "Existing" tells readers that no control was launched today. And "clarifications that the page applies to Google Search" signals a narrowing of scope rather than an expansion of it.

It is the fourth documentation change Google logged this month. On October 1, Google updated its guide on using generative AI content with material from the Search Quality Raters guidelines. On October 8 it added a page on the UGC Fresh Data Program. On October 16 it rewrote the Indexing API quota page, a change PPC Land covered when Google dropped the quota request form from the 200-a-day Indexing API page. None of those three entries touches the content controls page.

What the page looked like before

The changelog does not publish a diff, and the captured page offers no earlier version. Google's own translations of the page do. The Italian and German versions of the same URL, both still showing a last-updated date of December 31, 2025, list five methods: removing content from the site, password-protecting files, the noindex rule, disallowing crawling with robots.txt, and opting out of specific Google properties. Neither translation has a row for the robots meta tag group or for the generative AI control. Both mention data-nosnippet and X-Robots-Tag only in the left-hand navigation, as a link to a separate page.

The translations also show where the "Google Search" clarifications landed. The Italian page's main heading translates as "Control the content you share with Google," its table heading as "How to block content," and its final section as "Remove existing content from Google." In today's English version those read "Control what you share with Google Search," "How to block content from Google Search" and "Remove existing content from Google Search." Translations can lag or differ from the English source, so this comparison is indicative rather than exact. The left-hand navigation in the English capture still labels the page "Control what you share with Google," and the browser title is "Control the Content You Share on Search."

The seven methods

The table now runs to seven rows. Each pairs a method with an "Applicable" line describing what kind of content it works for. That second column is where most of the practical detail sits.

Remove the content from the site

"Applicable: all content types." Google calls deletion "the best way to ensure that it won't appear in Google Search and anywhere else on the Internet." It is the only method on the page whose effect extends beyond Google.

Password-protect files

Also "all content types." Restricting access to authorised users "will also prevent that content from appearing in Google Search, or if it already appears, it will eventually remove that content from our search results," according to the page. The word "eventually" is not quantified.

The noindex rule

"All content types" again. The page describes noindex as "a rule that tells Google not to index your content or let it appear in Google Search results." It also spells out the limit: content "can still be linked to and visited through other web pages, or directly visited by users with a link, but the content won't appear in Google Search results."

Disallow crawling with robots.txt

Here the scope narrows sharply. "Applicable: images and video." According to the page, "Google only indexes images and videos that Googlebot is allowed to crawl." The page does not present robots.txt as a way to keep web pages out of results, and that restriction is consistent with what Google has said for years. When Google's John Mueller discussed pages that were indexed despite a robots.txt block in September 2024, the explanation was that a crawler blocked by robots.txt never sees a noindex tag on the page, so external links can still put the URL in the index.

Robots meta tag, data-nosnippet or X-Robots-Tag

This is the first of the two new rows. Its applicability line is the least precise on the page: "various content types, depending on the implementation." The description reads: "You can use page- and text-level settings to adjust how Google presents your content in search results." The row links to Google's separate robots meta tag documentation for the details.

The overview page does not list individual rules such as nosnippet or max-snippet, and it does not say how any of them interact with AI features. Readers have to follow the link for that. The framing is also worth noting: this row is about how content is presented, not whether it is indexed. Noindex, technically a robots meta tag value too, keeps its own row higher up. The page does not explain the split, but it separates the rule that removes a page from the rules that shape a listing.

Those snippet rules have carried AI implications for some time. In March 2025, Google added AI Mode to its robots meta tag documentation, stating that nosnippet keeps content from being used as a direct input for AI Overviews and AI Mode. Until today, the overview page that site owners are pointed to for blocking options did not mention that family of controls at all.

The Search generative AI control

The second new row is the only one with "Applicable: site-wide." It reads in full: "You can manage your site's inclusion in generative AI features on Google Search, including AI Overviews and AI Mode, by using the Search generative AI control in Search Console."

That sentence names two features. The Search Console Help page it links to names three: AI Overviews, AI Mode and generative AI features in Google Discover. According to that Help page, the setting sits under Settings in Search Console and offers three choices: include, which is the default; exclude; and inherit from a parent property. Excluded content is no longer used as an input for generating AI responses, and the site receives no impressions or traffic from those features. Google says on the same Help page that the control is not used as a ranking signal elsewhere in Search, does not affect AI training, and does not override choices made in Merchant Center or Google Ads. Exclusion generally takes effect within one to two days, though caching can delay some content. The Help page dates the global rollout to August 31, 2026.

The "site-wide" label is the significant detail for publishers. The control works per property, not per page. PPC Land reported in June that page-level controls are tied to the UK Competition and Markets Authority's timetable, which requires them by March 3, 2027. Until then, a publisher wanting to remove some articles but not others from AI answers has the snippet rules in the row above, and the Search Console switch for everything at once.

Opt out of specific Google properties

The final row, "Applicable: web pages," lets site owners keep content out of named Google properties, "such as Google Shopping, Google Hotels, and vacation rentals." It links to a Search Console Help page on that opt-out. The page does not list AI features among those properties; they are handled by the row above.

What the page leaves out

A page that now "applies to Google Search" is, by definition, silent about Google's other uses of crawled content. The word Google-Extended does not appear anywhere in today's version. Neither does Gemini, Vertex AI or model training.

That omission is consistent with the stated scope, and the Search Console Help page for the generative AI control points readers to Google-Extended for training. But it means the most-linked overview of Google's content controls describes how to leave AI Overviews and AI Mode without mentioning the separate token that governs Gemini. Site owners who want to stay out of both still need two mechanisms, set in two places: a Search Console setting and a robots.txt rule.

The page is also silent on timing for most methods. Password protection removes content "eventually." The Help page gives one to two days for the AI control. No other row offers a figure.

The reasons Google gives for blocking

The page opens with three reasons a site might want content kept out of Search, and they have not changed in substance compared with the December 2025 translations. The first is restricting data meant only for visitors already on the site, with a reminder that files "may have metadata that can show up in Search." The second is low-value content. "Allowing indexing of such content may have a negative effect on your site's ranking in Google's search results," according to the page, which cites user-generated spam as the example. The third is crawl focus for very large sites, defined as "over hundreds of thousands of URLs," or sites with heavy duplication.

None of the three reasons mentions AI features. The page adds the AI control as a method, without adding traffic loss or content reuse as a reason to use it.

Why it matters for marketers and publishers

The update changes no behaviour, so its importance is navigational. Until today, the answers to publishers' questions about opting out of AI Overviews sat on pages other than this one. PPC Land reported in January that Google said it was exploring opt-out controls for AI search after the CMA proposed conduct requirements. In February, a Google executive called such a control a "huge engineering project," as PPC Land reported in its piece on Google saying publisher opt-outs from AI Overviews were a big engineering challenge. The toggle arrived on June 3 for a subset of sites, and by the end of August Google had rolled out AI reports and the control to every site.

Today's edit puts that control on the same table as noindex and robots.txt for the first time. That placement matters for anyone auditing a site, since agencies and in-house SEO teams often use this overview as a checklist. It also arrives seven weeks before December 3, 2026, when the main obligations of the CMA's Publisher Conduct Requirement take legal force in the UK.

Placement also invites comparison. Read top to bottom, the table now sets a whole-site AI switch beside rules that act on single pages or even single paragraphs. The trade-off is stark for a publisher that relies on Search traffic: data-nosnippet can hide a passage from snippets and AI inputs, while the Search Console control removes every page from AI Overviews and AI Mode at once, with the impressions that go with them. Google's own Help page says excluded sites get no traffic or impressions from those features.

What the documentation still does not do is quantify the cost of either choice. Search Console's generative AI performance report shows impressions but, as PPC Land noted in September, not clicks or queries. A site can now find every opt-out on one page. It still cannot see, from Google's own data, what clicks it would lose by using them.

Timeline

Summary

Who: Google, through its Search Central documentation, affecting site owners, publishers, SEO teams and agencies that manage how content appears in Google Search and its AI features.

What: Google updated the "Control what you share with Google Search" page to list seven methods for keeping content out of Search, adding rows for the robots meta tag, data-nosnippet and X-Robots-Tag, and for the site-wide Search generative AI control in Search Console, and clarified that the page applies to Google Search. Google says all the options already existed. The page does not mention Google-Extended or AI training.

When: Today, October 21, 2026, according to the Search Central changelog and the page's "Last updated" stamp. Earlier translations of the page date from December 31, 2025.

Where: On the Google Search Central documentation site, in the Removals section of the crawling and indexing documentation, with the AI control itself located in Search Console settings.

Why: According to Google, "to keep the page fresh and reflect the current set of options available to site owners." The update brings an AI opt-out that rolled out globally on August 31, 2026 onto the main overview of content controls, ahead of UK regulatory deadlines on December 3, 2026 and March 3, 2027.