The quick answer: open yoursite.com/robots.txt in your browser and look for GPTBot, ClaudeBot, PerplexityBot. If you see Disallow: / under one of these user-agents, that engine can't read the site. But this only checks technical access, not whether you're actually being cited — there are three separate levels.
Level 1 — Access: can the bot read the site?
Check robots.txt as above, by hand or with our free AI Crawler Access Checker (paste your domain, get the bot-by-bot list in seconds).
Level 2 — Structure: is the content easy to extract?
Even an accessible site can be hard for an LLM to "understand": long intros before the answer, no schema.org, no llms.txt. A full AIVIS scan checks all of this together, with a 0-100 score and the exact list of what's missing.
Level 3 — Citation: what the AI actually says about you, today
The only definitive test is to ask directly: our Brand Perception Check runs a real prompt against configured engines and shows the raw answer, not an inference.
A common mistake to avoid
Asking ChatGPT "do you know me?" once isn't a valid test: answers aren't deterministic and change over time. You need to track the same prompts periodically (see our article on how to measure brand citations).