Author: Krishna Madhavan, Principal Product Manager, Microsoft Bing Publishers should not have to choose between being discoverable in Bing and deciding how generative models use their work. Microsoft was the first company to provide publisher controls for AI training and grounding in 2023 with NoArchive and NoCache tags. Today, we are building on that commitment by offering newer, simpler, and clearer preferences that allow you to make choices about training, grounding, or both while remaining eligible for Bing search results. This announcement covers web publisher preferences for web content collected through our crawlers and used for training or to inform generated responses, including web-grounded responses in Bing, Copilot, and Microsoft 365 . Expanding web publisher choice and transparency Microsoft is working with other industry participants to bring a shared AI training preference to site owners and publishers. This step brings forward the year plus effort of the Internet Engineering Task Force (IETF) AI Preferences working group to achieve a global harmonized practice on AI preferences and is described in an Internet-Draft . This AI Training syntax gives web publishers a simpler, common way to express training preferences in robots.txt. The goal is to reduce the complexity of navigating different platform preferences for the same use of content. Although the efforts of the IETF AI Preferences Working Group continue toward a final standard, we believe the current work represents significant progress that merits adoption. We recognize additional work remains and Microsoft will continue to actively participate in this global standards effort. Microsoft is building on the shared AI Training preference and adding a separate AI Grounding preference, giving web publishers more granular, independent choices for training and grounding: AI Training is the shared preference expressed in robots.txt for a site or path. Microsoft is extending the same preference to the page level , so you can override the robots.txt setting for an individual page when needed. AI Grounding is a separate page-level preference from Microsoft that lets you manage how your content is used to help inform generated responses. Alongside Disa llow , which prevents grounding use, Microsoft is introducing a new granular preference: DisallowExtended . This limits grounding use to Microsoft-operated experiences such as AI Mode, AI Overview, Copilot, and Microsoft 365, and excludes grounding services Microsoft provides to third parties, such as Web IQ and grounding through API access in Foundry . If no grounding preference is expressed, your content may be used to help inform generated responses. These preferences will replace existing Microsoft preferences NoArchive and NoCache for training and grounding , allowing for greater flexibility and enhanced preference clarity. Starting February 1, 2027, Microsoft will honor the new preferences as it crawls: AI Training preferences in robots.txt apply to the relevant site or path, while page-level AI Training and AI Grounding preferences apply to individual pages. The new AI Training preference set at a page level overrides the robots.txt setting. If we detect new AI Training or AI Grounding preferences on a site, we will disregard any NoArchive and NoCache preferences that may remain on that site. These preferences provide you an ability to manage whether your content can be used for future AI generative model training and grounding experiences. If you select DisallowExtended or Disallow for grounding, you will see reduced brand visibility in AI experiences. Grounding connects generated responses to web content when an answer is created, bringing relevant sources and web publishers’ expertise into the response. Allowing grounding makes your content eligible to inform those responses, with links and citations providing a path to the original content. Allowing training can help improve future generative models. Neither choice guarantees inclusion, traffic, or payment. Act by March 1, 2027 The new preferences will be effective on February 1, 2027, as Microsoft crawls. If your site currently uses NoArchive or NoCache , we urge you to use these new preferences to manage whether your content should be used for training and grounding, and take the corresponding steps before March 1, 2027 : AI Training: Publish your AI Training preference in robots.txt or on individual pages. AI Grounding: Publish your AI Grounding preference on individual pages AND record your temporary transition grounding preference in Bing Webmaster Tools. For each page, the temporary preference applies until recrawl or September 30, 2027, whichever is earlier. To main tain your chosen preferences, complete the steps above by March 1, 2027; after that date, NoArchive and NoCache will no longer be read as a preference for training or grounding. The guide below explains the settings, transition safeguards, and important limitations. A companion FAQ is published in Bing Webmaster Tools help documentation . Implementation guide: important details and next steps The announcement above explains what is changing. This guide covers how to choose and publish your settings, and how the transition works. 1. Choose your training and grounding preferences independently Training and grounding are independent preferences. Choose the setting for each use below; your training setting does not change your grounding setting, and vice versa. Both default to Allow when no applicable preference is published. AI Training – Set in robots.txt or at the page level If your goal is to… AI Training preference Allow your content to be used to train generative models Allow – no action is necessary. It is the default when no applicable training preference is published. You can also explicitly set Allow. Prevent your content from being used to train generative models AI Grounding – page-level setting only If your goal is to… AI Grounding preference Allow your content to help inform generated responses Allow – no action is necessary. It is the default when no applicable grounding preference is published. You can also explicitly set Allow. Allow grounding only in Microsoft-operated experiences, but not through grounding services Microsoft provides to third parties DisallowExtended

Prevent your content from being used to help inform generated responses Figure 1. Independent training and grounding preferences for web publisher goals. For example, you can set AI Training to Disallow and AI Grounding to Allow to keep content eligible to inform generated responses without allowing it to train generative models. These usage preferences do not change eligibility for traditional Bing search results , subject to normal crawl and indexing requirements. They do not establish licensing or payment arrangements. 2. Understand what each preference covers AI Training covers using an asset to modify the learned parameters of a generative AI model . Under the IETF working definition of generative AI models, the training of models that are used to perform exclusively non-generative tasks is not included in this category, even if the model is capable of generative tasks. AI Grounding covers using content as input to help inform AI-generated responses. Indexing and ranking pages in traditional search results is not grounding. Set grounding preferences page by page, using one of these options: Allow : No action is necessary. It is the default. This means that content can be used to help inform generated responses. You can set if explicitly if you choose. DisallowExtended : Limit grounding use to Microsoft-operated products only: content can inform responses in experiences such as AI Mode, AI Overview, Copilot, and Microsoft 365, but not through grounding services Microsoft provides to third parties, such as Web IQ and grounding API in Foundry. Disallow : content is not used to help inform generated responses. The current Bing search syndication API returns links for traditional search scenarios, and is not intended for AI generated summaries, so it is not affected by the grounding preference. 3. Publish your settings by March 1, 2027 Training for a site or path. Use the shared robots.txt preference. The following examples show the available robots.txt patterns separately: a whole-site disallow, a path-level disallow, and an optional explicit whole-site allow. Allow is the default, so no robots.txt AI Training setting is needed to allow training. The following is an optional explicit Allow example: Training for a page. Add either a meta tag or an HTTP response header. These examples explicitly disallow training for the page: Grounding for a page. Add either a meta tag or an HTTP response header. These examples explicitly disallow grounding for the page: Use ai-grounding:disallowextended for the other grounding preferences. There is no site-wide robots.txt grounding preference. Caution: Selecting DisallowExtended or Disallow will significantly reduce your content’s visibility in AI-generated experiences, without affecting eligibility for Bing search results. Important configuration limitations: Both preferences default to Allow when no preference is set. Existing NoArchive and NoCache users should follow the transition steps below rather than assume their old restrictions carry over automatically. These settings govern use of content, not crawl access. An Allow setting does not override a robots.txt rule blocking Bingbot from crawling. Publishing a setting is not the same as activating it. Microsoft must crawl the page to discover the setting. IndexNow can signal that a page has changed and increase the chance of a recrawl, but it does not guarantee a recrawl date. 4. Plan the transition – applies if you use NoArchive or NoCache December 15, 2026 – temporary grounding transition preferences open in Bing Webmaster Tools: this transition applies only to and is only available for sites using NoArchive or NoCache as of that date. All other sites need no transition action. February 1, 2027 – new preferences begin taking effect: Microsoft starts honoring AI Training and AI Grounding preferences as pages are crawled. March 1, 2027 – deadline to maintain grounding restrictions: If you use NoArchive or NoCache and desire to withhold content from grounding, you must take the following steps before March 1, 2027 : (1) publish your AI Grounding preference on your pages; and (2) record a temporary grounding preference in Bing Webmaster Tools (BWT). We ask that you record your preference in BWT because we may not retrieve your new preferences until we recrawl your site; an activity that may not take place between February and March 1. September 30, 2027 – temporary preferences expire: a temporary grounding preference remains in effect until this date or until Microsoft recrawls the page and discovers its new AI Grounding preference, whichever occurs first. Training: publish your AI Training preference in robots.txt or on individual pages by March 1, 2027. After that date, NoArchive and NoCache will not be honored for training. N ew preferences take effect as Microsoft crawls, beginning February 1, 2027; there is no temporary training setting in Bing Webmaster Tools. Grounding: publish page preferences and use the temporary preference if your site uses NoArchive / NoCache as of December 15, 2026. The temporary Bing Webmaster Tools preference preserves your selected grounding treatment for pages awaiting recrawl. For each page, a new on-page AI Grounding preference replaces that temporary preference when Microsoft discovers it through recrawling. Any temporary preference still in place expires on September 30, 2027. If you do not record a temporary grounding preference by March 1 in Bing Webmaster Tools, pages using NoArchive or NoCache default to Allow for grounding until Microsoft recrawls them and discovers your new AI Grounding preference. Page-level s ettings published after March 1, 2027 can still take effect on recrawl, but a recrawl before the temporary preference expires is not guaranteed. Learn more: IETF Internet-Draft · Bing Webmaster Tools FAQ · Indexnow Related Stories New AI Visibility Insights in Bing Webmaster Tools: Intents, Topics, Citation Share, Compare Announcing Microsoft Web IQ A Smarter Way to Explore Images Has Come to Bing