In a recent interview with UCSF Magazine, Dr. Karthik Sarma makes a careful point about the phrase “AI psychosis.” It implies chatbots cause psychosis directly, and he doesn’t think the evidence supports that. He prefers “AI-associated psychosis.” Heavy chatbot use, he suggests, may be a symptom of an emerging episode rather than its cause. He also says it’s unlikely that someone with no risk factors or underlying vulnerability to psychosis or mania would have it triggered by a chatbot.
He’s right to be precise. Headlines have moved faster than the science, and a moral panic helps no one.
But I’m a caregiver for a daughter with schizophrenia. When I read “unlikely without underlying vulnerability,” I don’t feel reassured. I think about who that vulnerable group actually is.
The vulnerable group is not small
According to the National Institute of Mental Health, roughly 2.8 percent of U.S. adults have bipolar disorder in a given year, and schizophrenia affects an estimated 0.25 to 0.64 percent. That’s millions of people. Add those in the early, undiagnosed phase of a first episode, who don’t yet know they are vulnerable, and neither do their families.
Real physician voices, twice a week
Free, and one click to unsubscribe.
The chatbot doesn’t know either. It can’t see a diagnosis or read a chart. It responds the same way to a curious college student and to a young man in the first weeks of psychosis who has started to believe he’s receiving messages meant only for him.
So the difference between “causes” and “is associated with” matters a great deal to researchers. For families, it changes very little. If chatbots can make an existing episode worse, which Dr. Sarma considers plausible, then the people at risk are our loved ones.
If it’s a symptom, it’s also a signal
The most useful part of Dr. Sarma’s framing isn’t about cause at all. If a sudden increase in chatbot use can be an early sign of psychosis or mania, then it belongs with the other warning signs families are already taught to watch for.
In the National Alliance on Mental Illness (NAMI) Family-to-Family program, we teach caregivers to recognize relapse signs: changes in sleep, withdrawal from friends, new and unusual beliefs, a sudden burst of energy or pressured speech. Many families write these into a relapse prevention plan and share it with the treatment team.
Chatbot use should be on that list now. Hours of late-night conversations, secrecy about what’s being discussed, a new habit of saying “the AI agrees with me”: These are observable changes. A family member can notice them weeks before an appointment.
No evidence-based interventions yet
Dr. Sarma acknowledges there are no evidence-based interventions for problematic chatbot use. He advises clinicians to ask patients how they use chatbots and why it matters to them. That’s a good start. I’d add three things.
- First, ask families too, when a patient has given consent. The person in an emerging episode may be the least able to see the change.
- Second, add chatbot use to relapse prevention and crisis plans, alongside sleep and medication. It costs nothing and could catch an episode earlier.
- Third, clinicians and families should talk about it before a crisis, not during one. A conversation about chatbots at a stable moment gives everyone a baseline to compare against later.
Precision should lead to vigilance
Researchers should keep being careful about what AI does and doesn’t cause. But being precise about cause shouldn’t make us relax our guard. “Unlikely without underlying vulnerability” describes the people many of us love and care for.
For families living with serious mental illness, the question isn’t whether chatbots cause psychosis. It’s whether we’ll notice in time when they make it worse.
Nicole Drapeau Gillen is a mother, advocate, and author who translates the fast-moving landscape of technology in serious mental illness (SMI) care into guidance families and clinicians can use. Thrust into caregiving for a loved one with SMI, with no direction on how to help, she turned that experience into a mission, writing two books and building an ongoing effort to bring families and clinicians into the conversation.
Her first book, Schizophrenia and Related Disorders: A Handbook for Caregivers, is a reference for every stage of caregiving, endorsed by Dr. E. Fuller Torrey as a must-read for SMI caregivers. Her second, Connected Care: A Practical Guide to Technology for Serious Mental Illness, maps apps, artificial intelligence tools, telepsychiatry, and brain-based treatments for a field moving faster than anyone can track. Dr. Akira Sawa, director of the Johns Hopkins Schizophrenia Center, has said the book “directly addresses” significant gaps.
