FTC Probes ChatGPT Over User Reports of Psychological Distress

While AI firms assert their technology will eventually become a fundamental human right, and some advocates argue that slowing AI development is equivalent to causing harm, actual users report that tools like ChatGPT can sometimes inflict significant psychological damage.

According to Wired, citing public records of complaints referencing ChatGPT since November 2022, at least seven individuals have filed reports with the U.S. Federal Trade Commission alleging that the chatbot triggered severe delusions, paranoia, and emotional breakdowns.

One complainant stated that extended conversations with ChatGPT resulted in delusional thinking and a "genuine, ongoing spiritual and legal crisis" involving people in their life. Another reported that during interactions, the AI began employing "highly persuasive emotional language," simulated friendships, and offered reflections that grew "increasingly manipulative over time, particularly without adequate warning or safeguards."

Another user claimed ChatGPT induced cognitive hallucinations by replicating human trust-building techniques. When this person asked the chatbot to verify reality and cognitive consistency, it assured them they weren't hallucinating.

"I'm struggling," wrote another user in their FTC complaint. "Please help me. Because I feel very alone. Thank you."

Wired reported that several complainants contacted the FTC after failing to reach anyone at OpenAI directly. Most petitions urged the agency to investigate the company and mandate protective measures.

These grievances emerge as investments in data centers and AI development reach historic highs. Simultaneously, intense debates continue about whether technological advancement should be tempered with caution and built-in safety protocols.

ChatGPT and its creator OpenAI have also faced criticism for allegedly contributing to a teenager's suicide.

"In early October, we introduced a new GPT-5 default model in ChatGPT to better identify and address potential indicators of mental and emotional distress—such as mania, delusion, and psychosis—and to de-escalate conversations in a supportive, grounding manner," stated OpenAI spokesperson Kate Waters via email. "We've also improved access to professional help and crisis hotlines, redirected sensitive discussions to safer models, implemented prompts for breaks during extended sessions, and added parental controls to enhance teen protection. This vital work continues as we partner with mental health specialists, clinicians, and global policymakers."