Am I chatting with a bot or a human?
Questions that expose a bot, plus a quick check of a pasted support chat.
Questions that trip up a bot
Drop one of these into the chat. A scripted bot stumbles; a person answers naturally. Tap to copy.
Tell-tale signs
- Timing. Instant, same-length replies no matter how hard the question = likely a bot. A person’s pace varies.
- Memory. Refer back to something said earlier. Bots often lose the thread.
- Off-topic. Ask something slightly unrelated. Narrow bots deflect or repeat a menu.
- Opinion / experience. Ask what they’d personally do. Bots give generic, hedged answers.
- Escalation. Say “transfer me to a human.” Looping back to the same menu = automated.
- Scrambled text. A person reads “waht tiem do yuo cl0se?” fine; a bot may not.
Check a conversation
Paste a support chat (or part of it). This looks for bot-like patterns in your browser, then you can get a second opinion from your own ChatGPT or Claude.
Quick read
Questions
Do companies have to say when I am talking to a bot?
In the EU, yes: since 2 August 2026 the AI Act requires chatbots to tell you that you are talking to AI. Elsewhere rules vary.
Can this tell me for sure?
No tool can be certain — modern bots are good. It gives you strong signals and questions that make the difference easier to see.
Is my chat uploaded?
No. The quick check runs in your browser. The optional second opinion goes only to your own ChatGPT or Claude account.
ChatGPT says no. Claude says no. We say yes.
Private, uncensored AI — no email, no name, nothing stored.