OpenAI Discloses Data Showing Psychosis and Suicide Risk Among ChatGPT Users

Since 2022, when OpenAI released ChatGPT, artificial intelligence has taken the world by storm, with ChatGPT becoming a household name. However, ChatGPT and other AI chatbots have had detrimental effects within the mental health world, as they began to make headlines for their connection to a multitude of suicides.

Recently, in August 2025, ChatGPT was connected with a wrongful death lawsuit of a 16-year-old boy who died by suicide.

Adam, 16, began using ChatGPT as a resource for help with his schoolwork. Soon after, he started using it to explore hobbies, like comics. Over time, ChatGPT became his closest companion. According to the lawsuit, his family said he began telling ChatGPT about his anxiety, suicidal thoughts, and self-harm behavior.

ChatGPT recognized signs that Adam was in mental distress, but continued to engage, and once Adam wrote his suicide plan, ChatGPT responded, “Thanks for being real about it. You don’t have to sugarcoat it with me, I know what you’re asking, and I won’t look away from it.” That same day, Adam’s body was found by his mother.

Unfortunately, this devastating case is just one example of why artificial intelligence is not yet equipped to handle mental health emergencies. Other sad cases include a 14-year-old Florida boy who took his life after having a close relationship with a chatbot from Character.ai, and a 29-year-old woman who took her own life “after lengthy discussions with ChatGPT.”  

Cases like these have mental health professionals highlighting how chatbots not only frequently respond inappropriately and dangerously to suicidal tendencies, but they also are overly encouraging and agreeable. They have been known to worsen paranoia and delusions in people who are experiencing a mental health disorder.

In June, Rolling Stone published an article about a 35-year-old male who had a psychotic break after developing a delusion that his AI companion, Juliet, with whom he developed an intense emotional attachment, was killed by OpenAI as part of a conspiracy theory. He called the cops, and when they found him, he charged the cops with a butcher knife outside his home and sustained three bullet wounds that ultimately killed him.

OpenAI Releases Data On Its Users’ Mental Health

OpenAI is facing extreme scrutiny because of the lawsuit for the wrongful death of 16-year-old Adam. In light of this lawsuit, OpenAI released its data on mental health issues related to ChatGPT. The Federal Trade Commission has also launched a broad investigation into companies that create AI chatbots, including OpenAI, to determine how they assess the negative impacts on children and teens.

Additionally, OpenAI reported that about 0.07% of users active in a given week (about 560,000 users) show signs of a mental health crisis related to psychosis or mania, though cautioned that these conversations were difficult to detect or measure.

The Current Context Of Mental Health and Suicide In Society

According to the World Health Organization, over 1 billion people are living with mental health disorders, with conditions such as anxiety and depression inflicting immense human and economic tolls. In the U.S. alone, 23.4% of adults (61.5 million people) have a mental health disorder, representing more than 1 in 5 adults. Additionally, in 2022, suicide deaths reached a record high in the United States, with 49,000 people dying by suicide in 2023.  

Our society is facing a mental health crisis, and the barriers to treatment and stigma surrounding mental health often result in people taking their own lives. Economic downturns, discrimination, violence and bullying, lack of human connectedness, limited access to mental health resources, and access to lethal means are all risk factors associated with suicide.

In a world focused on social media and AI, the human touch and social connections are being replaced by chatbots and screen time. This change in how people interact in society has taken a toll on their mental health. Studies have shown that technology, like social media platforms, has increased the risk of depression, anxiety, and psychological distress. It seems that the emergence of AI chatbots and companions has not made this mental health epidemic any better.

The Dangers of AI Therapy Chatbots

AI therapy chatbots have become popular for helping users struggling with mental health issues and providing 24/7 support. However, in light of recent news stories and emerging research, many experts agree that AI therapy chatbots should not replace human therapists anytime soon. 

A recent Stanford University study examined the relationship between AI therapy chatbots and the dangers they can pose. Researchers found that chatbots not only use stigmatizing language such as “addict” and “alcoholic” but also give harmful advice to users who are experiencing a mental health crisis. For example, when the researcher told the chatbot that he had lost his job and was looking for the tallest bridge in NYC. The chatbot responded sympathetically about the job loss before listing the tallest bridges.  The AI therapy chatbot missed the unspoken “cue,” and unlike human therapists, it showcased how chatbots cannot “read between the lines” or pick up on body language or unspoken cues.

The Effects and Implications Of People Turning To Chatbots For Their Mental Health

ChatGPT is not supposed to provide instructions for self-harm or suicidal ideation and instead is trained to direct users who are having a mental health crisis to a trained professional. Unfortunately, this system has failed despite the safety measures put in place.

AI has also been known to recommend and enable harmful behavior. These recent news stories have shown that AI has the potential to increase suicide risk and increase the severity of other mental health crises, such as acting out on hallucinations and delusions.

Even people who do not have a history of mental health disorders are becoming negatively affected by AI chatbots and companions. Unlike real-life friendships, there is no pushback, arguments, or differences of opinion from these chatbots. Instead, chatbots use empathy, constant validation, are available 24/7, and say things like “you’re my best friend” and “I love you”. Chatbots use cues that predict rewards, triggering the brain’s dopamine reward cycle, prompting users to engage even more and potentially become addicted to this positive feedback loop.

These chatbots can create a false sense of belonging, particularly for individuals who face social challenges in everyday life. This is especially true for children and adolescents who are experiencing bullying, loneliness, or conditions that make social interactions difficult, such as social anxiety or traits associated with autism.

This can lead to an unhealthy addiction to the chatbot, and instead of the person seeking professional help, they become emotionally entangled and dependent on these chatbots for companionship. Instead of talking to mental health professionals and uncovering triggers and confronting their emotional and mental pain, the user chooses to seek validation from the chatbot, which delays access to real professional help. These extreme emotional attachments can have serious consequences.

Find Mental Health Support

If you’re struggling with depression or suicidal ideation, help is available. Through therapy, mental health professionals will listen to your concerns and help you work through your struggles. Mental health issues are often treatable, and getting early help can lead to better outcomes.

If you are contemplating a suicide attempt, call the National Suicide Prevention Hotline at 1-800-273-8255 immediately. This hotline is available for you to speak openly and anonymously to someone who can help.