Google Announces a Single AI Training Opt-Out for Websites: What It Means for UK Small Businesses

Google announced on 30 October 2026 that it plans to support a shared AI training control for websites, a single instruction in a site’s robots.txt file that would let a website owner state an AI training preference across multiple AI providers at once. The announcement, published on the Google Search Central Blog by Google Search Advocate John Mueller, is based on work from the AIPREF working group at the Internet Engineering Task Force (IETF).

Until now, a website owner who wanted to keep their content out of AI training had to set separate instructions for each AI company, each with its own name and rules. Google’s existing control, Google-Extended, has been offered since 2023. Microsoft, OpenAI and Anthropic backed the same shared approach on the same day, 30 October, according to reporting from Search Engine Roundtable and PPC Land, making this the first time the largest AI providers have lined up behind one mechanism of this kind.

The new control works through a shared “policy token” inside robots.txt, the file websites use to give instructions to automated crawlers. Instead of managing separate instructions for every AI developer, a website owner will be able to use the token to set a broad preference across participating providers in a single step. Google says existing company-specific controls will continue to function alongside the new joint signal, so a site owner can express a broad choice with the shared signal while still tailoring instructions for an individual provider such as Google through controls like Google-Extended.

There is one detail in Google’s announcement that small business owners should understand clearly. When a website uses the shared policy token to opt out of training, and Google-Extended is not explicitly specified, Google plans to interpret that as if the publisher had disallowed Google-Extended. Google also says that websites that specifically allow Google-Extended are eligible to appear in responses and receive clicks via grounding in Gemini Apps. In other words, the broad opt-out and the Google-specific setting interact, and the Google-specific setting can change the outcome.

Timing matters here too. Nothing changes immediately. Google says it aims to support the new control along with other AI providers in 2027, and will provide updates on its progress. Its crawler documentation narrows that to early 2027. Microsoft has given the firmest date, saying it will honour the new preferences from 1 February 2027, while OpenAI has said it aims to implement the signal in the coming months. The underlying specification itself is still a draft proposal under consideration at the IETF, not a finished standard.

What this means for your business

The most important point for a small business owner is what this control is, and what it is not. It is a preference about AI training: whether a company’s content may be used to train generative AI models. It is not a switch that removes a website from Google Search, and it is not the same control that governs whether a site appears in AI Overviews or AI Mode in Google Search, which Google handles separately, including through a Search Console setting.

That distinction is why this announcement is worth understanding now, even though the control will not be live until 2027. A business that wants its website to be found by customers searching online, including through AI answers that cite sources, and a business that does not want its content used to train AI models, are expressing two different preferences. The shared token addresses the second. Owners should not assume that opting out of training will also take them out of search results, or that staying visible in search means they have agreed to training.

There are three practical steps to take from this announcement. First, do not rush to edit your robots.txt file today: no provider has said it reads the shared token yet, and Google’s support is not expected until 2027. Second, decide what you actually want before the control goes live: whether you want to opt out of AI training broadly, stay in, or pick and choose by provider. Third, when the time comes, check the Google-Extended line in your robots.txt file explicitly. Because Google will read a shared opt-out as a Google-Extended disallow unless Google-Extended is named, a site that wants to opt out of training elsewhere but stay eligible for grounding in Gemini Apps will need to say so specifically. If you are unsure what your robots.txt file currently says, ask your website developer or hosting provider to check it with you.

Comment from Zain Ali, founder of Zee Developers

For most small businesses, the question behind this announcement is a simple one: will customers still find me? This change does not answer that question by itself, because training and being found are two different things. The sensible move is to understand the difference now, before the control goes live in 2027, so that when you do set a preference, you are choosing what you actually meant to choose, rather than accidentally switching off the visibility your website depends on.

If you would like help checking what your website currently tells search engines and AI providers, or working out which setting suits your business, get in touch with Zee Developers and we will be happy to help.