Is My Brand In AI — AI Bot Checker Review (2026): Free, No Signup, No-Log Fetch

What the Is My Brand In AI bot checker does

Type a domain into the AI Bot Checker and you get back one narrow, load-bearing answer: which AI crawlers your robots.txt currently turns away. The checker lives on ismybrandinai.com, an independent AI-search-visibility and GEO (generative engine optimization) research platform whose organizing question is whether your brand shows up when people ask ChatGPT, Perplexity, Gemini or Google's AI Overviews.

A robots.txt line is the cheapest possible way to lose that game. One inherited Disallow, written years ago with scrapers in mind, quietly removes your site from the sources ChatGPT is allowed to read, and nothing warns you. Your pages keep ranking in classic search while the answer engines assemble their responses from your competitors. This checker exists to surface that silent misconfiguration in the time it takes to paste a domain.

Which bots it checks

Coverage has grown since this review first ran. When we published in June 2026, the tool grouped its checks under six engine families; the version of the page we re-read on August 27, 2026 lists eight, several of them carrying more agents than before:

• ChatGPT (OpenAI): GPTBot, OAI-SearchBot, ChatGPT-User
• Claude (Anthropic): ClaudeBot, anthropic-ai, Claude-Web
• Perplexity: PerplexityBot, Perplexity-User
• Gemini: Google-Extended
• Apple Intelligence: Applebot-Extended
• Common Crawl: CCBot
• Meta AI: meta-externalagent, FacebookBot
• ByteDance: Bytespider

Three of those tokens do not appear in the vendors' own current documentation, and two that do are missing. We read both vendor pages on August 28, 2026. Anthropic's crawler page documents exactly three agents (ClaudeBot, Claude-User and Claude-SearchBot), and anthropic-ai and Claude-Web appear nowhere on it. Anthropic does not call them deprecated; they are simply absent from the page it maintains, which is weaker evidence than a retirement notice and is the strongest evidence available. Meta's crawler page lists FacebookExternalHit, Meta-WebIndexer, Meta-ExternalAds, Meta-ExternalAgent and Meta-ExternalFetcher, and FacebookBot is not among them.

This matters in a specific way for anyone copying the list into a file. A robots.txt built from this tool's Claude family blocks two agents Anthropic no longer documents while leaving Claude-User and Claude-SearchBot (the ones that fetch pages when somebody asks Claude a question) unblocked and unmentioned. The tool would report your site as covered for Claude. It would be reporting on a rule that current Claude traffic does not match. That is a gap in the shortlist, not in your syntax, and it is the risk that comes with curation: a hand-picked list is only as fresh as its last curation pass.

The readout stays granular rather than binary. A crawler shows as allowed (with the page distinguishing "no disallow rules" from "specific paths disallowed") or as blocked, which the tool defines as a Disallow: / that no Allow: / overrides. A site with no robots.txt at all gets its own state ("nothing is blocked"), which is the correct reading of an absent file.

Curation is still the design philosophy, even at eight families. BrandCited will read the same file against roughly 64 user-agents; this tool picks the crawler families with a plausible path to putting your brand inside an answer and skips the rest. Notably, the expansion closed most of the gaps we flagged in June (CCBot, Bytespider and Meta's agents are now on the list), so today's trade-off is less "major omissions" and more "shortlist versus census."

How we reviewed this

This review is documentary. Everything below about the checker's behavior is drawn from the tool's own page and the published material on the wider ismybrandinai.com platform, read against the AI companies' own published user-agent names.

What the tool's own page documents:

• It is free, with no signup, no card and no email gate: you paste a domain and get a result.
• Its page spells out the fetch path: a small server-side function retrieves the target site's robots.txt and returns the text, and the parsing happens in your browser. On the domain field, the page's own words are "No account, no log, no database." The retention section below unpacks this.
• As the tool documents, it reads the site's robots.txt and reports a per-bot allow/block/partly-limited status for each of the bots it covers, rather than collapsing them into a single pass/fail.
• It surfaces the honest GEO note (see below) rather than implying a clean result means you're winning.

We do not run hands-on tests, we conducted no accuracy benchmark, and we report no accuracy percentage of our own, nor did we fabricate a metric or a screenshot to stand in for one. Where we describe limits (server-side blocks, the long-tail gap), those are reasoned from how the tool works by its own description, not from a lab measurement.

Disclosure

AI Tools Police earns affiliate commissions on some tools we cover. This is not one of them. The Is My Brand In AI bot checker is free, has no paid signup, and has no affiliate program. We make nothing if you use it, and we're ranking it first anyway. The affiliateSlug in this page's metadata is routing plumbing for our internal links; no commission sits behind it. We sell no GEO platform of our own, which is exactly why we can call a free tool the category's best pick without flinching.

What happens to the domain you type in

For a tool whose entire interface is a text box, "where does the text go?" is the trust question. So for this update we re-read the tool page and the platform's privacy policy (both fetched August 27, 2026) specifically to answer it.

Start with a correction of our own earlier framing, finished on August 29, 2026. We previously described the checker as running entirely in your browser, and when we corrected that here, the page's own title and privacy score kept the old claim: the title still said "In-Browser" and the privacy dimension still scored a flat 5.0. Both are fixed now. A perfect privacy score would mean there is no privacy cost, and there is a small one: the vendor's server sees which domain you asked about, and the assurance that it keeps no record of it is the vendor's own published statement rather than something anyone outside can check. The no-signup half of that dimension really is faultless; the privacy half rests on a claim, so the dimension now reads 4.5. The tool's current page is more precise, and slightly different: a small function on the vendor's side fetches the target site's robots.txt and returns the text, and "your browser does the parsing." That server assist is technically necessary (a browser generally cannot pull another site's robots.txt cross-origin), and the page is equally direct about what rides along with it: the domain you type "is passed to the fetch and nothing else," with "No account, no log, no database" behind the field.

The platform's privacy policy is consistent with that minimal footprint. It states the site sets no cookies at all (no advertising, no tracking) and uses Cloudflare Web Analytics, which it notes stores no personal data. The only personal information the policy claims to hold is an email address you volunteer through downloads or the contact form, deletable on request. The policy doesn't discuss tool inputs at all, which is consistent with the tool page's position that no input is stored in the first place, though it means the no-log promise rests on the tool page alone.

Two honest qualifiers. These are the vendor's published statements, not an audit anyone outside the vendor can perform. A no-log claim is inherently unverifiable from the outside. And the precision cuts in the tool's favor: a page that documents its own fetch function in plain language is making a narrower, more falsifiable promise than a vague "100% private" badge, which is part of why this tool holds our transparency-led top spot.

The honest part: necessary, but not sufficient

What ultimately separates this checker from the pack is that its own page refuses to oversell a green result. The current version compresses the asymmetry into seven words:

Blocking guarantees invisibility, but unblocking guarantees nothing.

Both halves earn their place. The first half is absolute: a blocked crawler cannot read you, so a blocking rule is a guaranteed loss, the one outcome a robots.txt fix reliably prevents. The second half is the part vendors usually mumble: an open file grants permission, and permission is inert. Nothing about an allowed GPTBot obliges OpenAI's systems to fetch your pages, weigh them, or ever surface your name; that gets decided downstream, by what your pages say and by how often the wider web already mentions your brand. (The platform's own research leans on the finding that unlinked brand mentions track AI citations more closely than backlinks do.)

Run the checker, in other words, to eliminate the guaranteed loss. What remains afterward is the unguaranteed game, and a tool that says so on its own landing page is doing right by you in a category where plenty of competitors let a page of checkmarks imply victory.

Where it stops being enough

The checker's fetch function sees exactly what a polite crawler sees: the contents of one file. That single fact defines both of its ceilings.

Bot scope is the softer one. Eight families cover every mainstream answer engine, but the long tail of training scrapers and niche agents keeps growing, and anything outside the list (Amazonbot, for instance) simply isn't examined. If your job calls for a full census of who may touch your content, BrandCited's ~64-agent sweep exists for exactly that.

The harder ceiling is that robots.txt is a statement of policy, not the enforcement layer. Real enforcement often lives in infrastructure the file never mentions: a CDN bot-management preset, a firewall rule keyed on user-agent strings, an IP-reputation block at the edge. Those act before your origin serves anything, so they are structurally invisible to any robots.txt parser, this one included. A crawler can be welcomed by your file and still turned away at your CDN, and no result here will tell you. So read the output as "my stated policy is correct," verify the enforcement layer in your own infrastructure, and re-run the check whenever the file changes, since nothing here watches it for you.

Verdict

Is My Brand In AI's AI Bot Checker is our top pick among the AI bot checkers we cover, and it earns that on transparency, privacy and ease, not on raw bot count. It's free, it asks for nothing, it documents its own no-log fetch path instead of hand-waving about privacy, and it prints the caveat most vendors bury: unblocking guarantees nothing. The trade-off remains coverage: eight curated crawler families rather than BrandCited's ~64-agent census, and a view that ends at robots.txt rather than your server config. If you want a fast, private, no-nonsense check of whether the AI engines that matter can actually reach your site, start here; if your job needs the full user-agent census, that's BrandCited's lane. The whole field, ranked: our best AI bot checkers roundup.