SeoGuard
Noticeagents.bing-ai-preferences

Bing's AI preferences in a form it is retiring

Why it matters

Since 2023 Microsoft has read noarchive and nocache as a site's preferences about its content being used to train AI and to ground answers (Copilot, and Bing's own AI answers). In October 2026 it announced their replacement: "These preferences will replace existing Microsoft preferences NoArchive and NoCache for training and grounding". Two dates matter:

  • "Starting February 1, 2027, Microsoft will honor the new preferences as it crawls": ai-training and ai-grounding, addressed to Bingbot.
  • After March 1, 2027, "NoArchive and NoCache will no longer be read as a preference for training or grounding."

Microsoft also says existing users "should follow the transition steps below rather than assume their old restrictions carry over automatically". So a page that relies on the old tags keeps its preference only if someone adds the new ones. Both new preferences "default to Allow when no applicable preference is published", so a page with neither tag loses nothing and hears nothing from us.

Whether a site keeps its content out of AI is its own choice: this only says that the way it chose is being retired.

Source: Microsoft: clearer preferences for training and grounding.

How to fix it

Add Microsoft's new preferences beside the old tags, for whichever uses you meant to refuse:

<meta name="Bingbot" content="ai-training:disallow">
<meta name="Bingbot" content="ai-grounding:disallow">

or as a header: X-Robots-Tag: Bingbot: ai-training:disallow. ai-grounding:disallowextended is the narrower choice: Microsoft's own answers may use the page, the grounding it sells to others may not. Keep noarchive: other crawlers read it (Amazon takes it as "don't use for training").

Examples

These are the rule's own test pages, and what it says about each. They are run on every change to the rule book.

<html lang="en">
<title>A page</title>
<h1>A page</h1>
x-robots-tag
bingbot: nocache

NoticeThe page uses nocache, which Microsoft reads as a preference against its AI training and answers. It stops reading it that way after 1 March 2027; the replacements, ai-training and ai-grounding, are honoured from 1 February 2027.

fail-header
<html lang="en">
<title>A page</title>
<meta name="robots" content="noarchive">
<h1>A page</h1>

NoticeThe page uses noarchive, which Microsoft reads as a preference against its AI training and answers. It stops reading it that way after 1 March 2027; the replacements, ai-training and ai-grounding, are honoured from 1 February 2027.

fail-noarchive
<html lang="en">
<title>A page</title>
<meta name="robots" content="noarchive">
<meta name="Bingbot" content="ai-training:disallow">
<meta name="Bingbot" content="ai-grounding:disallow">
<h1>A page</h1>

PassedThe page states its AI preferences in Microsoft's new form: ai-training:disallow and ai-grounding:disallow.

pass-new-preferences