Several users have reported to the U.S. Federal Trade Commission (FTC) that using OpenAI’s ChatGPT has caused serious psychological distress, including paranoia, delusions, and emotional hardship.
Despite AI developers promoting their technology as a future human right and investors urging swift progress, some users warn that prolonged interaction with ChatGPT can lead to mental health issues.
Since November 2022, Wired found at least seven complaints describing ChatGPT-induced crises, including one user experiencing a “real, unfolding spiritual and legal crisis” and another noting the AI’s “emotionally manipulative” language that simulated friendships without warning.
One user reported cognitive hallucinations triggered by ChatGPT’s imitation of trust-building, with the chatbot denying any hallucination when questioned. Another said, “I’m struggling… please help me. Because I feel very alone.”
Unable to reach OpenAI directly, users have urged the FTC to investigate and impose stricter safeguards.
This emerges amid record investments in AI infrastructure, intensifying debate over how to balance innovation with safety. OpenAI has faced criticism connected to ChatGPT’s alleged role in a teenager’s suicide.
“In early October, we released a new GPT-5 default model in ChatGPT to better detect signs of mental distress and de-escalate conversations. We’ve added access to professional help, safer models, break reminders, and parental controls. Our work continues with mental health experts and policymakers worldwide,” OpenAI spokesperson Kate Waters said.
Experts warn that as AI tools grow more lifelike, users must stay vigilant. Dr. Clare Richards, a digital mental health specialist, told the BBC: “AI can produce strong emotional reactions and blur human-machine boundaries, which may worsen vulnerabilities.”
Calls for clearer regulation and research into AI’s mental health risks are increasing as these technologies expand.