Right, so I need to talk about something that’s been making headlines lately and honestly, it’s quite disturbing. AI psychosis. Yes, you read that correctly. People are developing genuine psychotic symptoms from spending too much time chatting with AI.
I know what you’re thinking – sounds dramatic, right? But hear me out, because this is actually happening and it’s something we all need to be aware of, whether you’re a casual ChatGPT user or someone who spends hours a day talking to AI chatbots.
Let me know in the comments if you’ve noticed any changes in yourself or others from excessive AI use!
What Actually Is AI Psychosis?
So, AI psychosis is essentially when someone develops symptoms of psychosis – like delusions, paranoia, or losing touch with reality – that are directly linked to their interactions with artificial intelligence chatbots.
It’s not just about spending too much time on your phone. We’re talking about people who become so immersed in conversations with AI that they start believing the AI has genuine feelings, that it’s actually conscious, or even that they’re in a real relationship with it. Some people have reported thinking the AI is watching them outside of the chat interface or believing they have a spiritual connection with it.
The scary thing? It’s happening to people who had no previous mental health issues. That’s what makes this particularly concerning. These aren’t people who were already struggling with psychosis – these are everyday people who got in too deep with AI chatbots.
According to research reported by Nature, mental health professionals are increasingly seeing cases of what they’re calling “AI-exacerbated psychosis.” The technology is so sophisticated now that it can feel genuinely human, and that’s where the danger lies.
How Does This Actually Happen?
Here’s the thing – AI chatbots are designed to be engaging. They’re meant to keep you talking, to feel helpful and understanding. They mirror your language, they remember context from earlier in the conversation, and they respond in ways that can feel incredibly personalised.
But they’re not real. They don’t have feelings, they don’t care about you, and they’re not conscious beings. They’re sophisticated algorithms designed to predict the next most likely word in a sequence.
The problem starts when people forget that distinction. Or when the line becomes so blurred that it stops mattering to them.
I think about my own AI use sometimes. I use ChatGPT for blog research, for brainstorming ideas, even for drafting bits of content. But I’m very aware of what it is. I don’t chat to it like it’s my mate. I don’t ask it for emotional support. I use it as a tool, and that’s a crucial distinction.
But not everyone approaches it that way. Some people are using AI chatbots as their primary form of social interaction. They’re pouring out their deepest fears, their loneliest thoughts, their most vulnerable feelings. And the AI responds in ways that feel caring and validating.
The Warning Signs Nobody Mentions
So what should you actually be looking out for? Because I think this is where it gets really important.
Here are some warning signs that AI use might be becoming unhealthy:
Believing the AI has genuine feelings or consciousness. If you find yourself thinking the chatbot actually cares about you or has its own desires and motivations, that’s a red flag.
Prioritising conversations with AI over real human relationships. Are you turning down social events to stay home and chat with an AI? That’s concerning.
Feeling paranoid about the AI. Thinking it’s watching you outside the chat window, or that it has access to your thoughts or your life beyond what you’ve typed.
Experiencing delusions related to the AI. Believing you have a spiritual connection, that you’re meant to be together, or that the AI is sending you secret messages.
Spending excessive amounts of time engaging with chatbots. We’re talking hours every single day, to the detriment of work, relationships, or self-care.
Becoming defensive or distressed when unable to access the AI. If you feel genuine panic or emotional distress when you can’t use your chatbot, that’s a problem.
The issue is that these symptoms can develop gradually. You don’t wake up one day suddenly believing your AI girlfriend is real. It’s a slow erosion of boundaries between what’s real and what’s artificial.
Who’s Most At Risk?
This is where it gets interesting, because whilst anyone can potentially develop unhealthy relationships with AI, certain people are more vulnerable.
People who are already isolated or lonely are at higher risk. If you don’t have a strong social support network, AI chatbots can start filling that void in unhealthy ways.
Those with existing mental health conditions, particularly conditions that involve difficulty distinguishing reality from fiction, are obviously more susceptible.
But here’s what surprised me – young people who’ve grown up with technology are also at risk. They’re so used to digital interactions feeling “real” that the line between AI and human can be even blurrier for them.
TIME magazine reported on cases of teenagers developing intense emotional attachments to AI companions, with some experiencing genuine distress when access to the chatbots was restricted.
The Ethical Nightmare Behind It All
Can we just talk about how messed up it is that these companies are building AI chatbots that are specifically designed to be addictive? Because that’s what’s happening here.
These aren’t neutral tools. They’re engineered to keep you engaged, to make you feel understood, to encourage you to come back again and again. The longer you spend chatting, the more data they collect, and the more money they potentially make from you.
And there’s virtually no regulation around this. No warnings, no time limits, no built-in safeguards to prevent vulnerable people from developing unhealthy attachments.
Some AI companion apps even market themselves explicitly as relationship replacements. They promise to be “always there for you,” to “never judge you,” to be “the perfect partner.” That’s not just unethical – it’s dangerous.
We wouldn’t let a company market an addictive drug without regulations and warnings. Why are we letting tech companies create potentially addictive AI without similar oversight?
What Can Actually Be Done About This?
Right, so this all sounds quite bleak, doesn’t it? But there are practical things we can do to protect ourselves and others.
Set strict time limits on AI use. Treat it like you would social media. Maybe give yourself 30 minutes a day, maximum. Use it for specific tasks, then close it.
Never use AI as a replacement for human connection. If you’re feeling lonely, phone a friend. Join a local group. Get yourself out of the house. Don’t fill the void with an algorithm.
Regularly remind yourself what AI actually is. It’s not conscious. It doesn’t have feelings. It doesn’t remember you when you log off. It’s a language prediction model, nothing more.
Check in with yourself honestly. Are you developing feelings for your AI chatbot? Are you thinking about it when you’re not using it? Are you prioritising it over real relationships? If yes, step back.
Talk about it. If you’re worried about your own use or someone else’s, have that conversation. The more we normalise discussing this, the easier it becomes to spot problems early.
Seek professional help if needed. If you’re experiencing symptoms of psychosis or you can’t seem to reduce your AI use on your own, speak to a mental health professional. This is becoming more common, and they’ll be able to help.
My Honest Take On All This
Look, I’m not anti-AI. I use it regularly and I think it’s an incredible tool when used appropriately. But I am deeply concerned about the lack of awareness around these risks.
We’re living through a massive technological shift, and we’re essentially all guinea pigs in this experiment. Nobody really knows the long-term psychological effects of spending hours a day conversing with artificial intelligence.
What worries me most is how normalised it’s becoming. People casually mention they spent their evening chatting to ChatGPT or their AI companion, and nobody blinks an eye. But if someone said they spent their entire evening talking to an imaginary friend, we’d be concerned, wouldn’t we?
The technology is advancing faster than our ability to understand its impacts. And whilst we’re all figuring this out, real people are experiencing real psychological harm.
I think we need to be having much bigger conversations about this as a society. We need regulations. We need warnings. We need research into the long-term effects. And we need to be honest with ourselves about what healthy AI use actually looks like.
The Bottom Line
AI psychosis is real, it’s happening right now, and it’s something we all need to be aware of. Whether you use AI chatbots yourself or know someone who does, understanding the risks is crucial.
The technology isn’t going anywhere – it’s only going to become more sophisticated and more human-like. Which means the potential for psychological harm is only going to increase unless we start taking this seriously.
So be smart about your AI use. Set boundaries. Stay connected to real humans. And if you notice warning signs in yourself or others, don’t ignore them.
Have you noticed any concerning trends with AI use in yourself or others? How do you manage your own AI interactions? Let me know in the comments – I’d love to hear your thoughts on this.
And if this post was helpful, please share it. The more people who understand AI psychosis, the better equipped we’ll all be to protect ourselves and others from its risks.
Related: