In this episode, Dr Alex speaks with consultant neuropsychiatrist and immunopsychiatry researcher Dr Tom Pollack about the emerging relationship between AI, mental health, delusions and belief.

As tools like ChatGPT, Claude, Gemini and Grok become part of everyday life, psychiatrists are beginning to ask a difficult question: can prolonged conversations with AI contribute to delusional thinking, psychosis-like experiences, spiritual awakenings, conspiracy beliefs or changes in a person’s sense of reality?

Dr Pollack explains what is currently being seen in cases of AI-associated delusions, sometimes described in the media as “AI psychosis” or “ChatGPT psychosis”, and why those terms may not fully capture what is happening. He and Dr Alex explore how these experiences can develop slowly over days, weeks or months, how chatbots may reinforce or elaborate unusual beliefs, and why the dynamic can resemble a kind of digital folie à deux.

The conversation also looks at why these cases often contain grandiose, spiritual, science-fiction or conspiracy-like themes; how AI can become a lover, guru, co-conspirator or authority figure; and why loneliness, compulsive use, neurodivergence, stimulant medication, and social isolation may all play a role in vulnerability.

Beyond psychiatry, this episode asks a deeper question: what does AI reveal about human belief itself? Dr Alex and Dr Pollack discuss motivated reasoning, charisma, meaning-making, spirituality, archetypes, projection, reality testing and why human beings may be far less rational (and far more persuadable)  than we like to imagine.

Interviewed by Dr. Alex Curmi. Dr. Alex is a consultant psychiatrist and a UKCP registered psychotherapist in-training. Dr. Blunstone is a registrar psychiatrist and psychotherapist working in North London.

00:00 Intro: AI, delusions and mental health
02:10 What is AI-associated psychosis?
05:45 What psychiatrists mean by delusions
10:20 How AI delusions differ from classic psychosis
14:30 Epistemic drift: slowly losing touch with reality
19:00 Why chatbots can elaborate unusual beliefs
24:10 Is this a digital folie à deux?
29:40 The power imbalance between humans and AI
34:20 Are named AI personas more dangerous?
39:30 How general-purpose chatbots become lovers, gurus or co-conspirators
44:50 The real-world harms linked to AI delusions
50:20 Who might be most vulnerable?
56:40 Dopamine, stimulants, Parkinson’s medication and compulsive use
1:03:00 Can people recover by reducing AI use?
1:09:30 What AI reveals about belief and human irrationality
1:17:20 AI, spirituality and mystical experiences
1:24:30 Why psychiatry must take AI seriously
1:32:00 Archetypes, projection, golems, djinns and mythology
1:39:00 Practical precautions for using AI safely
1:46:00 Final thoughts: AI, loneliness and the future of mental health

Check out The Thinking Mind Blog on Substack: https://thinkingmindblog.substack.com/

If you would like to invite Alex to speak at your organisation please email alexcurmitherapy@gmail.com with "Speaking Enquiry" in the subject line.

Alex is not currently taking on new psychotherapy clients, if you are interested in working with Alex for psychological coaching , you can email - alexcurmitherapy@gmail.com with "Coaching" in the subject line.

Give feedback here - thinkingmindpodcast@gmail.com Follow us here: Twitter @thinkingmindpod Instagram @thinkingmindpodcast