Microsoft fair announced new Bing AI satisfied controls "offering newer, simpler, and clearer preferences that authorize you to create choices concerning training, grounding, or the two during remaining eligible for Bing hunt results," Krishna Madhavan, Principal Product Manager, Microsoft Bing, wrote. Google is also on board, it seems.
In summary, this gives location owners three chief options, including:
- AI Training environment to authorize or forestall your satisfied from being used to train AI models.
- AI Grounding environment to authorize or forestall your satisfied from being used to assistance create AI answers.
- Limited Grounding environment to authorize your satisfied to assistance Microsoft's own AI products, but not its third-party grounding services.
This method you can obstacle AI training but motionless appear in AI answers. For example, a publishing company could forestall Microsoft from using its articles to train models during motionless allowing Copilot to citation those articles. You can obstacle AI answers from using your satisfied but doing so can decrease your visibility in AI experiences. The traditional Bing rankings are unaffected.
Here is the timeline for these changes:
- December 15, 2026: Temporary grounding controls rotate into accessible in Bing Webmaster Tools for eligible sites using the older NoArchive or NoCache directives.
- February 1, 2027: Microsoft starts honoring the new preferences as it crawls pages.
- March 1, 2027: Deadline for publishers using the old directives to configure the new preferences if they desire to keep their restrictions.
- September 30, 2027: Temporary grounding preferences expire.
The changes appear to add on to Bing's existing controls for managing how website satisfied appears in AI-generated answers. Bing Webmaster Tools additionally offers an AI Performance reporting that shows citations, cited URLs and grounding queries throughout supported AI experiences. The new preferences provision publishers additional granular authority complete how their satisfied can be used, separating AI example training from the use of satisfied to create AI answers during preserving traditional hunt visibility.
There are a lot additional item over current in the blog post and the documentation complete here.
Microsoft is joining industry peers in announcing a shared AI training penchant in robots.txt - and going additional alongside granular page-level training preferences and a distinct grounding penchant for publishers and location owners, the business announced.
Forum conversation at X.
