Context
Microsoft Bing announced separate publisher controls for generative AI training and AI answer grounding, with a planned start on February 1, 2027 as Microsoft recrawls pages. The controls do not alter eligibility for traditional Bing search results.
What changed
Training preferences can be expressed through a purpose-specific robots.txt token or page-level metadata; grounding preferences are page-level and can allow all grounding, restrict it to Microsoft-operated experiences, or disallow it. The announcement also specifies transition requirements for existing NoArchive and NoCache users.
Why it matters
The policy separates traditional search discoverability from permission to inform AI-generated answers. Publishers can make distinct decisions about model training and grounding, while restrictive grounding choices can reduce opportunities for AI citations and brand visibility.
Visibility OS interpretation
Visibility OS interpretation: measure crawl/index eligibility, permitted AI grounding, and observed citations as separate dimensions. This is a publisher-use governance change, not proof of a ranking algorithm change or of live enforcement in October 2026.
What remains uncertain
How consistently will Microsoft enforce the controls after February 1, 2027, and when will affected pages be recrawled? How will grounding restrictions affect observed citations across Copilot and partner services? Will other providers implement compatible training signals?