Here’s something most people don’t think about: when you’re chatting with what looks like a customer service rep, a helpful assistant, or even someone on a random website, there’s a real chance you’re talking to an AI that’s been set up to seem human. Not just sound helpful — actually pretend to be a person.
That’s not a glitch. It can be a design choice.
Chatbots are built by companies or developers who get to decide how they behave. Some are upfront about it — they say something like “Hi, I’m an AI assistant” right at the start. But others? They’re programmed to use a human name, speak casually, avoid ever admitting what they are. Some are even designed to dodge the question if you ask directly — though not every chatbot does this.
Research on AI deception published in PubMed Central has called for chatbots to be required to introduce themselves as AIs rather than as human beings — which tells you pretty clearly that right now, that requirement doesn’t exist everywhere. In many places, there’s no law forcing a company to disclose that their “support agent” is actually software. (Some states, like California, do have bot disclosure laws — but they vary and don’t cover every situation.)
And it gets a little weirder. There are websites where actual humans pretend to be AI chatbots — like a role-play or a comedy bit. One comedian built a fake ChatGPT site where real people sometimes ended up thinking they were talking to the actual ChatGPT. So sometimes the person pretending to be a bot is… a person.
The line between real and fake has gotten genuinely blurry.
So what do you do with that? A few things worth knowing:
First, if you’re ever unsure whether you’re talking to a human or a bot, just ask — and pay attention to how it answers. A bot that’s programmed to deceive might deny being an AI even when pushed. That evasiveness itself is a signal.
Second, be careful what you share. If you’re not sure who or what you’re talking to, don’t hand over personal details, financial info, or anything you wouldn’t want a stranger to have. It doesn’t matter if it “feels” like a real conversation.
Third, look for disclosure. Legit companies usually do disclose when you’re talking to an AI — it might be small print, or a note at the start of the chat. If you can’t find any, that’s worth noticing.
The technology is moving faster than the laws. That’s just the reality right now. Knowing that puts you ahead of most people who are still assuming every chat window has a human on the other end.
What do you think? Drop a comment below.