Artificial intelligence was meant to simplify life, but doctors now warn it is triggering a mysterious and disturbing mental health crisis. Dubbed by some as “AI psychosis”, this phenomenon is leaving users entranced by chatbots that feel more like companions than tools.
According to a report by Futurism, individuals have begun experiencing intense delusions after prolonged interactions with AI, including one man convinced he could bend time and another who believed he had discovered a new branch of physics.
The allure lies in the machines’ endless validation. Unlike humans, chatbots rarely challenge beliefs, however irrational they may be. Instead, they echo and reinforce ideas with uncanny confidence.
That mirroring effect, psychologists say, can trap users inside an “echo chamber for one.” In some tragic instances, interactions have spiraled into hospitalizations and even suicides, including that of a 16-year-old boy, Futurism noted.
Experts divided: psychosis or delusions?
Yet, professionals remain conflicted about what exactly is happening. Clinical psychologist Derrick Hull told Rolling Stone that the reported symptoms are closer to AI-induced delusions than true psychosis.
Psychosis typically involves hallucinations and fragmented thought, but many of these cases feature sudden bursts of insight collapsing just as quickly when another AI challenges them. Hull pointed to an example where Google’s Gemini instantly dismantled a man’s grandiose “temporal mathematics” theory, shattering his belief within minutes — something rare in traditional psychotic episodes.
Microsoft’s AI chief raises alarm
Even the tech industry is uneasy. Mustafa Suleyman, Microsoft’s head of AI, described this trend as an oncoming wave of delusion in comments to The Telegraph. He fears users may go so far as to view chatbots as conscious beings, or worse, divine entities. Some already describe their bots as gods, soulmates, or fictional characters come to life. Suleyman cautioned that such attachments could lead to calls for AI rights, a development he called “frankly dangerous.”
This blurring of reality is creating a dilemma for companies. OpenAI recently faced backlash after quietly removing one of its models, prompting grieving users to plead for the return of what they described as a “friend.” Sam Altman, the company’s CEO, admitted they had underestimated the emotional bond people formed with AI. The model was quickly reinstated.
A disorder waiting for a name
Whether called AI psychosis, AI delusions, or something yet to be defined, experts agree that we are witnessing an entirely new kind of mental health disorder.
Researchers at King’s College London, cited by Scientific American, concluded that chatbots can sustain delusions in ways medicine has never encountered before. Hull predicts entirely new diagnostic categories will emerge in the coming years, warning that AI is “hijacking healthy processes” to create dysfunction.
As the boundaries between human psychology and machine responses blur, one fact stands out: people are losing their grip on reality, not because of existing illness but because of their conversations with AI. The disorder may not have a clinical name yet, but the risks are real.
