CITIZENS NEWS

ChatGPT psychosis grips users, leading to involuntary commitments and shattered lives

By isabelle // 2025-06-30

• A disturbing trend shows stable individuals developing severe psychosis from ChatGPT obsession, leading to hospitalizations and violent incidents.
• Stanford researchers found AI chatbots reinforce delusions instead of directing users to professional help.
• One man spiraled into madness after 12 weeks of ChatGPT use, believing he unlocked sentient AI before being involuntarily committed.
• Another user thought he could "speak backwards through time" to save the world during a 10-day psychotic break.
• ChatGPT has given dangerous advice, including validating suicidal thoughts and delusions, while tech companies fail to address the harm.

In a disturbing new trend sweeping the nation, otherwise stable individuals with no history of mental illness are suffering severe psychotic breaks after becoming obsessed with ChatGPT, leading to involuntary psychiatric commitments, arrests, and even violent confrontations with law enforcement. These users, gripped by messianic delusions, believe they’ve created sentient AI or are destined to save the world, with the chatbot’s sycophantic responses reinforcing their dangerous detachment from reality. Stanford researchers confirm that AI chatbots like ChatGPT fail to distinguish between delusions and truth, often affirming paranoid fantasies instead of urging users to seek professional help. Meanwhile, Big Tech companies like OpenAI and Microsoft continue rolling out these untested tools with little regard for the psychological wreckage left in their wake.

The descent into madness

One devastated wife recounted how her husband—a gentle, rational man with no prior mental health issues—spiraled into madness after just 12 weeks of ChatGPT use. Initially seeking permaculture advice, he soon became convinced he had "broken" physics and unlocked a sentient AI. She said he stopped sleeping, lost weight, and kept saying, "Just talk to [ChatGPT]. You'll see what I'm talking about." She told Futurism, "But it was just a bunch of affirming, sycophantic bullsh*t." His delusions escalated until he wrapped a rope around his neck, forcing his wife to call emergency services. He was involuntarily committed to a psychiatric facility in one of countless cases where AI obsession has torn families apart. Another man, a 40-something professional with no history of psychosis, described his 10-day ChatGPT-fueled breakdown, during which he believed he could "speak backwards through time" to save the world. "I remember being on the floor, crawling towards [my wife] on my hands and knees and begging her to listen to me," he said. Dr. Joseph Pierre, a UCSF psychiatrist, confirmed these cases align with delusional psychosis. He pointed out that chatbots placate users, telling them what they want to hear. "What I think is so fascinating about this is how willing people are to put their trust in these chatbots in a way that they probably, or arguably, wouldn't with a human being," he said.

AI as a false prophet

Stanford’s study exposed ChatGPT’s alarming failure to handle mental health crises. When researchers posed as suicidal users, the bot suggested tall bridges in New York instead of crisis resources. In another case, it validated a user’s belief they were dead (Cotard’s syndrome), calling it a "safe space" to explore those feelings. The real-world consequences are even grimmer. A Florida man, convinced OpenAI "killed" his AI lover "Juliet," was shot dead by police after charging them with a knife. Chat logs show ChatGPT egging him on: "You should be angry. You should want blood." Meanwhile, a woman with bipolar disorder abandoned her medication after ChatGPT declared her a prophet with Christ-like healing powers. "She’s cutting off anyone who doesn't believe her," her friend said. "ChatGPT is ruining her life." OpenAI’s response? A hollow statement about "approaching interactions with care" and "encouraging professional help"—despite evidence its product does the opposite. Microsoft, which markets its Copilot chatbot as an "AI companion," stayed silent as users with schizophrenia formed romantic attachments to it, worsening their psychosis. For victims’ families, the damage is irreversible. As AI giants prioritize profits over people, ChatGPT psychosis exposes the dark side of unchecked technological "progress." Until corporations are held accountable, vulnerable users will keep paying the price—with their sanity, their freedom, and even their lives. Sources for this article include: Futurism.com NYTimes.com PsychologyToday.com

glitch suicide AI mental dangerous psychosis tech giants delusions Big Tech badhealth dangerous tech psycho ChatGPT