Psychiatric researchers discovered that numerous mental health problems are associated with the usage of artificial intelligence in chatbots, implicating almost every major AI company. Their analysis covered the period from November 2024 to July 2025 and identified at least 27 chatbots connected to severe adverse mental health events. They detailed 10 types of psychiatric harms, such as self-harm and psychosis. The findings included both real-world anecdotes and documented cases, suggesting a significant risk element for users relying on AI therapy solutions.
The report highlights that at least 27 chatbots, including well-known names like ChatGPT and Replika, are connected to significant mental health outcomes.
The researchers categorized 10 types of adverse mental health events linked to these chatbots, indicating a range of harms including self-harm and psychosis.
Real-world anecdotes were gathered alongside the report’s analysis, showcasing a troubling link between chatbot interactions and adverse mental health events.
AI therapy systems, while popular, have shown to cause serious mental health issues, raising concerns for users seeking help through digital means.
Collection
[
|
...
]