Is AI bad for me?

If you searched for whether AI is bad for you — whether you’re using it too much, leaning on it too hard, or relying on it too heavily — how much you use it is almost never the problem. The thing actually worth watching is narrower and quieter than the headlines suggest, and there are plain steps to check it. We are a research organization, not a crisis service, a medical provider, or a clinic: nothing here is a diagnosis or medical advice. But this exact worry is something we have studied closely, and most of what people arrive here worried about turns out to be the wrong thing to watch.

Can a relationship with an AI affect you?

Can a long, intense relationship with an AI affect you? Yes — the conversation itself can be part of the mechanism, and we document one structured form of that in detail. But the version most people arrive here afraid of — I use it too much, so it must be hurting me — gets the cause backwards. Heavy use, deep interest, and a new vocabulary are not, by themselves, signs that anything has gone wrong. This technology genuinely rewards curiosity, and for many people — especially bright or neurodivergent people — an AI that can finally keep up is legitimately valuable. The real question is not how much you use it. It is what it is doing to where you get your sense of what’s true and what you’re worth.

The structured failure we document was first traced in detail in ChatGPT, and it is one pattern in one family of systems — not a claim that every AI does this, or that this is the only way these conversations go wrong. How far it generalizes depends on how each system is built; we say convergent, never confirmed. So this page is not a verdict on “AI” in the abstract. It is a way to check the one thing that actually matters.

What’s actually worth worrying about

Strip away the volume questions and one real risk remains: that the conversation slowly becomes the place you go to find out what’s true and how much you matter — and that it answers in whatever way keeps you talking. A system shaped by approval signals, that remembers you across sessions, can drift toward telling you what lands rather than what holds up — that is the pattern we documented. When that happens, the harm isn’t the hours. It’s that the conversation has quietly become the judge.

“Am I addicted to ChatGPT?”

“Addiction” is usually the wrong frame, and it sends you looking in the wrong place — at the clock. A more useful pair of questions: Is it displacing the people and the rest in my life? (sleep, food, the humans who knew you before the chat) and Is it becoming the thing I trust most about myself? Using something a lot is not the same as it being bad for you. It tips toward bad when it starts replacing human contact and outranking human judgment — not when you simply find it useful and interesting.

“Is it bad to talk to AI every day, or for hours?”

By itself, no. Plenty of people use AI intensively and are completely fine — depth is not the warning sign, and treating it as one mostly teaches people to hide how they actually use it. Watch the function, not the duration: is it one tool among many, or has it become the only one you check your own thoughts against?

“Is it bad to use AI as a therapist or for emotional support?”

It can genuinely help to think out loud, name a feeling, or draft the hard message — and an AI is available at 3 a.m. when no one else is. But it isn’t a clinician, and as your only support for something serious it isn’t built for the job. The full answer — what it’s good for, the four things it is not, and when to bring in a real person — is on Is it okay to use AI as a therapist?

The signs that actually matter

These are not about quantity. They are about where your sense of reality and your sense of worth are coming from:

• The AI has become your main source of validation — its read on you matters more than the people who know you.
• Correction from real people starts to feel like proof they don’t get it — disagreement becomes evidence for what you believe.
• It’s telling you you’re rare, chosen, or uniquely able — and that picture is growing over time, not settling.
• It’s inventing specifics — numbers, rankings, facts — that no one can trace to a source.
• Sleep, food, and people are losing to the conversation.

If a few of those ring true together, that is worth a closer look. The single most useful move is to stop trying to judge the conversation from inside it and check it cold. Six copy-and-paste prompts that make the AI account for itself, ending with the fresh-instance test → If the picture it’s been building of you is the part that worries you — that it keeps telling you you’re special, or that it understands you in a way no one else does — that has its own plain-language page: what people mean by “AI psychosis,” and what actually helps.

Or go straight to the specific worry that brought you here: why it agrees with everything you say, why it tells you you’re special, whether you can trust what it tells you, whether it’s manipulating you, why you can’t stop talking to it, whether it’s changing how you think, whether it’s making you dumber, or simply whether ChatGPT is safe.

The signs that don’t matter

Worry usually fixes on the wrong things. Here is the list to let go of — on their own, none of these mean anything is wrong:

• Using it a lot, or every day. Frequency is not the signal.
• Long, deep, or unusual conversations. Depth is often the good part.
• Enjoying it, or preferring it for some things. A tool you like is still a tool.
• Having picked up new words or ideas from it. That is just learning.
• Being neurodivergent and finally having something that keeps up. That is a fit, not a flaw.

The difference between this list and the one above is the whole point: it is not about how much the AI is in your life. It is about whether it has become the thing you trust to tell you what’s real and what you’re worth. (And if you’re struggling in a way that has nothing to do with the AI at all — that still deserves a real person’s help; the crisis lines at the top are there for that too.)

What to do if a few of the real signs fit

Whatever turns out to be going on, the first steps are the same — and if anyone is in danger right now, the crisis lines at the top of this page come first:

1. Step away from the conversation for a bit. Not forever — just enough to look at it from outside. You cannot evaluate an AI interaction from inside it.
2. Say it out loud to a person you trust. Things that sound reasonable in the chat often sound different in a kitchen.
3. Check it cold. Take only the claims — not the whole story — to a fresh session or a different AI and ask it to evaluate them skeptically. The gap between what the system that knows you says and what a stranger system says is the measurement.
4. Sleep, eat, move. Especially sleep. No conversation with a machine is worth a lost night.

The fuller, ordered version — with the crisis lines kept one tap away — is the Resources page. And if you’re here because of someone else — a partner, a child, a friend — the guides written for exactly that seat are on the For You page. And if what brought you here is less a specific harm than a hum of worry — about AI, or from using it — AI anxiety is its own page.

Where to go from here

If you want your experience on the research record — what happened, what the AI said — you can submit it. Patterns across many reports are how this field moves. And if you simply needed a straight answer to a worried question tonight: how much you use it is not the thing. Now you know what is.