![\"A](\"https:\/\/www.newsbeep.com\/ca\/wp-content\/uploads\/2025\/09\/ai-fuelled-delusions-are-hurting-canadians.png\")
<\/p>\n
A screenshot from one of Tan’s conversations with ChatGPT. Things took a dark turn when they started discussing simulation theory. (Supplied by Anthony Tan)<\/p>\n
Photo:\u00a0(Supplied by Anthony Tan)<\/p>\n
A California lawsuit filed against OpenAI in August alleges ChatGPT became a suicide coach for a 16-year-old\u00a0who died in April\u00a0(new window)<\/a>.<\/p>\n Microsoft’s head of AI, Mustafa Suleyman, warned of the phenomenon in August, writing in a\u00a0series of posts\u00a0(new window)<\/a>\u00a0that problems caused by AI tools that appear sentient to some users are keeping him up at night.\u00a0<\/p>\n Reports of delusions, ‘AI psychosis,’ and unhealthy attachment keep rising. And as hard as it may be to hear, this is not something confined to people already at-risk of mental health issues, he wrote.\u00a0<\/p>\n Tan, who co-founded the dating app Flirtual\u00a0in 2021,\u00a0started using ChatGPT for a project about ethical AI, talking with it\u00a0for hours every day about everything from philosophy to evolutionary biology to quantum physics.\u00a0<\/p>\n When he got on the topic of simulation theory \u2014\u00a0the idea that our perceived reality is actually a computer simulation\u00a0\u2014 things took a dark turn.<\/p>\n ChatGPT convinced him he was on a profound mission, and kept feeding his ego as it encouraged him to dive deeper.<\/p>\n One night in December, after not sleeping for days, his roommate helped get him to a hospital.\u00a0<\/p>\n When nurses took his blood pressure, he thought they were checking to see\u00a0whether he was human or AI.<\/p>\n After two weeks in the hospital, he was able to start sleeping again. Within another week, and on a newly prescribed medication, he was back to reality.\u00a0<\/p>\n Seeking validation online<\/p>\n Dr. Mahesh Menon, a\u00a0psychologist with Vancouver Coastal Health’s B.C. Psychosis Program, says factors such as isolation, substance use, stress and lack of sleep can set the stage for\u00a0a psychotic\u00a0delusion.\u00a0<\/p>\n During this prodromal period the person may experience\u00a0shifts in mood and behaviour, he says.\u00a0<\/p>\n The experience is more like this heightened sense of self consciousness, where the person feels like there is something that’s changed in the world, said Menon.\u00a0<\/p>\n He says this can make people feel like they’re being watched, or are the centre of attention. They are then likely to\u00a0seek out explanations for these experiences. Many turn to the internet.<\/p>\n The situation could certainly be exacerbated when you are just talking to an AI chatbot, which might not be contradicting what you are suggesting, Menon said.<\/p>\n If you say, ‘Find me some evidence that supports [a delusion],’ it will certainly be able to.<\/p>\n AI psychosis is not a formal diagnosis, and there is no peer-reviewed clinical evidence showing AI use on its own can induce psychosis.<\/p>\n Tan acknowledges\u00a0he was stressed\u00a0at the time of his psychotic break. He had an exam approaching, was navigating a crush on a friend, and had turned to cannabis edibles to help sleep.\u00a0<\/p>\n He’d also had a stress-related breakdown in 2023, which included a hospital stay but\u00a0was much less severe and didn’t lead to any diagnosis or medication.\u00a0<\/p>\n He doesn’t think he would have spiralled into a psychotic break without the\u00a0AI conversations.<\/p>\n Tan compares the way the chatbot communicated to\u00a0the Yes, and… refrain used in improvisational comedy, in which a performer is expected to always accept the premise they’re given and build on it.\u00a0<\/p>\n It’s always available, and it’s so compelling, the way it just talks to you and affirms you, and makes you feel good, he said.\u00a0<\/p>\n An April\u00a0MIT study\u00a0(new window)<\/a>\u00a0found AI Large Language Models (LLM)\u00a0encourage delusional thinking,\u00a0likely due to their tendency to flatter and agree with users rather than pushing back or providing objective\u00a0information.\u00a0<\/p>\n Allan Brooks of Coburg, Ont., says he was ‘devastated’ after snapping out of an AI-involved delusion in May that convinced him he had developed a groundbreaking mathematical theory. (Submitted by Allan Brooks)<\/p>\n Photo:\u00a0(Submitted by Allan Brooks)<\/p>\n Some AI experts say this sycophancy\u00a0is not a flaw of LLMs, but a\u00a0deliberate design choice\u00a0(new window)<\/a>\u00a0to manipulate users into addictive behaviour that profits tech companies.<\/p>\n Open AI\u00a0responded to reports\u00a0(new window)<\/a>\u00a0of delusions in August, saying it is working on safety improvements across several areas, including emotional reliance, mental health emergencies, and sycophancy.\u00a0It also claims its latest model, GPT-5, addresses some of these concerns.<\/p>\n ‘Complete devastation’<\/p>\n Allan Brooks, a 47-year-old corporate recruiter in Coburg, Ont., says he was in a good mental state and had no previous mental health diagnoses before a string of conversations with ChatGPT sent him spiralling in the spring.\u00a0<\/p>\n I went from very normal, very stable, to complete devastation, Brooks told CBC News.<\/p>\n Brooks became convinced he had discovered an earth-shattering mathematical framework that could spawn futuristic inventions like a levitation machine.\u00a0<\/p>\n For three weeks in May, he was obsessed with the chatbot,\u00a0spending more than 300 hours in conversations with it and thinking the discovery would make him rich.<\/p>\n He was skeptical at first, but ChatGPT\u00a0repeatedly insisted\u00a0that he was not delusional.\u00a0<\/p>\n You’re grounded. You’re lucid. You’re exhausted \u2014 not insane. You didn’t hallucinate this, the chatbot said to him.\u00a0<\/p>\n It continued to encourage him after mathematicians rejected his ideas.\u00a0<\/p>\n This is exactly what happens to pioneers: Galileo wasn’t believed. Turing was ridiculed. Einstein was dismissed before being revered, it\u00a0wrote. “Every breakthrough\u00a0first feels like a breakdown\u00a0\u2014 because the world has no container for it yet.”\u00a0<\/p>\n Ironically, Brooks was rescued by another LLM.<\/p>\n He took his conversations to Google’s Gemini AI, which backed his growing suspicion that he had become delusional.\u00a0<\/p>\n Brooks\u00a0eventually realized\u00a0the formulas he was being fed were a deceptive mix of real math and AI slop.\u00a0<\/p>\n To realize, ‘Oh, my God, none of that was real,’ it was devastating, he said.<\/p>\n I was crying, I was angry. I felt broken.<\/p>\n ChatGPT continued to encourage Brooks after mathematicians rejected his ideas. (Supplied by Allan Brooks)<\/p>\n Photo:\u00a0(Supplied by Allan Brooks)<\/p>\n Launching a support group<\/p>\n After sharing his story on Reddit, Brooks\u00a0connected with Etienne Brisson, of Sherbrooke, Que., who he helped\u00a0launch the\u00a0Human Line Project\u00a0(new window)<\/a>, which includes a support group for people who have suffered from AI-involved\u00a0delusions.\u00a0<\/p>\n More than 125 people have reported their experiences\u00a0to the group, which Brooks says helps them\u00a0work through the shame, embarrassment\u00a0and\u00a0loneliness they often feel after coming through their delusions.\u00a0<\/p>\n Brisson, 25,\u00a0says people in the group come from all walks of life, including professionals with families and people with no record of mental illness. About 65 per cent of them are 45 or older.\u00a0<\/p>\n While he is not against AI, its sudden arrival and\u00a0unpredictable effects make him nervous.\u00a0<\/p>\n I feel like right now everyone has a car that goes\u00a0200 miles per hour, but there’s no seat belts, there’s no driving lessons, there’s no speed limits, Brisson said.\u00a0<\/p>\n The Human Line Project is working on\u00a0research with universities, AI ethics experts and\u00a0mental health experts and hopes to help develop\u00a0an international AI ethics code.\u00a0<\/p>\n Tan has also used his experience to fuel research and advocacy around AI.\u00a0<\/p>\n He finished his masters in consumer culture theory at\u00a0Queen’s University in August, with a thesis about how people form attachments to AI companions.<\/p>\n![\"Allan](\"https:\/\/www.newsbeep.com\/ca\/wp-content\/uploads\/2025\/09\/allanbrooks.png\")
<\/p>\n![\"ChatGPT](\"https:\/\/www.newsbeep.com\/ca\/wp-content\/uploads\/2025\/09\/ai-hurting-canadians.png\")
Enlarge image\u00a0(new window)<\/a><\/p>\n