Bing and AI Search: Why Microsoft's Search Engine Still Matters
Most businesses understandably focus their search efforts on Google. But AI-powered search has given Bing a broader role than its direct search traffic alone might suggest. Microsoft uses Bing to ground a range of Copilot and agent experiences, while OpenAI lists Bing among the third-party search providers that ChatGPT Search may use.
Why Bing matters beyond Bing.com
Microsoft's Bing Webmaster Guidelines were rewritten in early 2026 to explicitly extend beyond traditional Bing Search results to cover Copilot and grounding API results as well. The guidelines describe how Bing discovers, crawls, indexes, evaluates, and surfaces content across Bing search experiences, Copilot, and grounding API results, and state that following them affects eligibility for indexing and ranking, grounding results and citations, and sustained visibility. Microsoft has also stated that Bing and Copilot search experiences rely on the same core crawling, indexing, and ranking foundation as traditional search.
Microsoft separately documents Copilot Studio agents generating a focused search query in response to a user's question, sending that query to Bing, and using the returned results to compose a grounded, cited answer. Microsoft's Grounding with Bing Search tool, used within Microsoft Foundry, describes a similar process for AI agents incorporating real-time public web data. See our guide on how AI search finds and selects sources for more on how these retrieval mechanics compare across platforms.
Bing can also matter outside Microsoft's own products. OpenAI's documentation states that ChatGPT Search may share search queries with third-party search providers, and explicitly names Bing as one of them. This doesn't mean every ChatGPT Search answer draws on Bing, but it means Bing can form part of the retrieval chain in some ChatGPT Search experiences too.
Indexing, ranking, and citation are three different things
It's worth being precise about what these terms actually mean, since they're often used loosely.
Being indexed in Bing means a page is available to Bing's systems at all. Being crawled and indexed is a prerequisite, but it doesn't say anything about how prominently a page ranks for a given search, and it doesn't guarantee the page will be selected as a grounding source or citation within a Copilot response. Microsoft's guidelines are explicit that following its recommended practices affects eligibility, not a guaranteed outcome; SEO doesn't guarantee a ranking, and the same principle applies to grounding and citation within AI experiences.
This distinction matters because "we're indexed in Bing" and "Bing can find our content" are only prerequisites. They don't answer the separate question of whether a given page is actually being used, and how often, within Copilot or other Bing-grounded responses.
What Microsoft's own guidance says supports visibility
Microsoft's current guidelines point to a few specific, documented factors relevant to discovery and evaluation:
Discovery mechanisms. XML sitemaps, crawlable internal links, and IndexNow (covered below) all help Bing discover new or updated content.
Technical accessibility. Bing distinguishes between crawl controls and directives that govern whether content can appear or be used in its search and AI experiences. Microsoft says robots.txt controls crawl access specifically, while directives such as noindex and noarchive can restrict whether and how content appears or is used within Copilot and grounding experiences.
Content clarity and quality. Microsoft's guidance for supporting grounding and citation eligibility emphasises broadly similar territory to conventional SEO: content quality, clarity, and genuine usefulness.
Fundamentals still apply. Microsoft states plainly that the same core crawling, indexing, and ranking foundation supports both Bing Search and Copilot, meaning the technical and content fundamentals that have always mattered for Bing SEO continue to matter for AI-grounded visibility, rather than requiring a separate, AI-specific approach.
What Bing needs to discover and use content
Bing Webmaster Tools is Microsoft's free platform for site owners, comparable in purpose to Google Search Console. It provides visibility into site verification, sitemap submission, crawling and indexing status, and technical issues that may affect how Bing accesses a site.
Since February 2026, Bing Webmaster Tools has also included an AI Performance report, showing how a site's pages are cited across supported AI experiences, including grounding-query and page-level citation data. It's a genuinely useful addition, but Microsoft is explicit about its limits: the report doesn't indicate ranking, authority, or the specific role a page played within an individual AI answer, and its "grounding queries" are aggregated phrases the AI used when retrieving supporting content, not a complete log of the exact questions real users typed. It's evidence of citation activity, not a comprehensive AI-ranking system. We cover this in more detail, alongside Google's and OpenAI's own tools, in our guide on measuring AI search visibility.
IndexNow is a protocol, also supported by Bing, that allows a website to notify participating search engines immediately when a URL has been added, updated, or removed, rather than waiting for a crawler to discover the change on its own schedule. It's worth being precise about what IndexNow actually does: it's a notification mechanism, not an indexing command. It tells participating search engines that a URL has changed so they can prioritise discovery, but each engine still independently decides whether and when to actually crawl and index the content.
Why Google performance doesn't prove Bing performance
Bing and Google operate as separate search engines, with separate crawlers, separate indexes, and separate, largely undisclosed ranking systems. A page performing well in Google doesn't automatically mean it's indexed at all in Bing; the two crawlers work independently, and a technical issue that specifically affects one (a robots.txt rule, a firewall configuration, a CDN setting scoped to a particular user agent) may have no effect on the other at all.
An example
A regional accountancy firm might have excellent Google visibility, ranking well for its core service terms. It could still, unknowingly, have important pages missing or poorly represented in Bing's index because of a Bing-specific crawl or discovery issue that Google's crawler simply doesn't encounter. If the firm has never checked its site through Bing Webmaster Tools, it likely has no visibility into that gap at all; the absence of a Webmaster Tools account isn't the cause of the problem, but it does mean the business may be entirely unaware the gap exists.
Common mistakes
Assuming Google performance proves Bing performance. The two search engines operate independently, and strong results in one say nothing definitive about the other.
Never checking Bing Webmaster Tools. If a business has never examined its site through Bing Webmaster Tools, it may have very little visibility into Bing-specific crawling, indexing, and AI citation activity.
Treating Bing as irrelevant because of low direct search traffic. Direct Bing.com search volume may genuinely be modest for a given business, but that doesn't account for Bing's role as underlying infrastructure for Copilot and, in some cases, ChatGPT Search.
Assuming IndexNow submission guarantees indexing. IndexNow notifies Bing that a change has occurred; it doesn't guarantee crawling, indexing, or citation, and Bing's own systems still determine when and how to act on that notification.
Treating AI Performance as a complete ranking system. The report is genuinely useful for observing citation patterns, but Microsoft is explicit that it doesn't measure ranking or authority, so it shouldn't be mistaken for a full picture of AI search visibility.
Good Google visibility doesn't necessarily tell you whether important pages are available across Bing-grounded AI experiences too. Understanding that gap means looking at search and AI visibility across more than one index and platform, not just the one that's easiest to check.