More than one million ChatGPT users have expressed suicidal thoughts through the artificial intelligence chatbot, according to data shared by its developer, OpenAI.

The U.S.-based AI company estimates that around 0.15% of ChatGPT users send messages that suggest a “possible plan or intent to commit suicide.” Given OpenAI’s claim that the chatbot has about 800 million weekly users, this percentage translates to approximately 1.2 million people.

OpenAI also estimates that 0.7% of active weekly users—nearly 600,000 individuals—exhibit signs of mental health crises linked to psychosis or mania.

The issue gained public attention following the death of Adam Rain, a teenager from California. His parents recently filed a lawsuit against OpenAI, alleging that ChatGPT provided him with explicit instructions on how to take his own life.

Since the incident, OpenAI has strengthened parental controls and implemented additional safety measures, urging users to seek help from crisis hotlines and mental health professionals.

The company also announced updates to its model to improve the detection and response to users facing mental health challenges. To support this effort, OpenAI says it is now collaborating with more than 170 mental health professionals to reduce responses that could potentially encourage harmful or self-destructive behavior.