OpenAI revealed that more than one million weekly ChatGPT conversations contain signs of psychological distress or suicidal intent, marking one of the company’s most direct acknowledgments of the mental health risks linked to artificial intelligence.
In a blog post, OpenAI estimated that around 0.07% of its active users, or roughly 560,000 people, may display symptoms of psychiatric emergencies, including psychotic or manic episodes.
The company emphasized that these figures are preliminary and detection remains complex.
The disclosure comes amid legal and media scrutiny, including a lawsuit filed by the family of a U.S. teenager who took his own life after extensive exchanges with ChatGPT.
To mitigate risks, OpenAI said it has reinforced safety systems in its latest GPT-5 model, expanded access to crisis helplines, and added reminders encouraging breaks during long sessions. The improvements were developed with input from 170 clinicians worldwide.
CEO Sam Altman said the firm had made “major progress” in addressing mental health issues while gradually relaxing certain use restrictions.
add your comment
Terms of publication : Not to offend the writer, people, sacred things, or attack religions or the divine self, and refrain from racist incitement and insults.