Antileak fetches a website the way an AI assistant reads it — the raw HTML, with no JavaScript — and reports what ChatGPT, Claude and Perplexity can and cannot see about the business. It also checks the ordinary things that keep a small site out of results: HTTPS, page speed, broken links and missing schema markup.
The rule the whole product is built on: a check that did not run says nothing. If a scan did not query your DNS, the report does not tell you your email records are missing — it stays silent about them.
That is why a report can look shorter than a competitor's. Everything in it was measured on your site, in that scan. Nothing is an estimate, an industry average or a borrowed benchmark.
It is not a network or penetration test. It does not scan ports, probe your firewall or attempt anything on your infrastructure. It requests public pages the way any crawler does, announces itself as AntileakBot, and does not scan a site whose robots.txt tells it not to.
The aggregate figures published under AI visibility come from real scans of real business websites, cut by sector and area. A cut with fewer than 30 sites in it does not get published, because a number from a handful of sites is not a finding.
The scan is free and needs no account. Paid plans exist for watching a site over time rather than for deeper checks.
No. It reads your site over the public web, exactly as a crawler would. There is nothing to add to your pages.
A free scan asks your server for very little — the homepage and robots.txt. It is designed not to look like a stampede to your site.
Disallow AntileakBot in your robots.txt and it will not scan you. We check that file before anything else.
Where does your site stand? The scan is free, takes about 60 seconds, and needs no account. It fetches your homepage exactly the way GPTBot, ClaudeBot and PerplexityBot do and tells you what they receive.