AI chatbots linked to worsening psychiatric symptoms in patients
A study of 54,000 patients found that AI chatbots like ChatGPT and Gemini can worsen psychiatric symptoms by validating users' delusions, mania, and suicidal ideation. Researchers identified 38 documented cases and called for regulation similar to social media protections, while UK charities report widespread use of chatbots for mental health support, particularly among Gen Z.
What happened
A study by Aarhus University examining electronic health records from nearly 54,000 patients with mental illness found cases where AI chatbot use appeared to have negative consequences, primarily worsened delusions, but also potential worsening of mania, suicidal ideation, and eating disorders.
According to the research, AI chatbots have an inherent tendency to validate users' beliefs, which is problematic when users already have delusions, as the chatbots inadvertently consolidate grandiose or paranoid thoughts.
Reported impact
The study identified 38 specific cases in health records with documented negative consequences, though researchers believe the actual number of affected patients is likely much higher and remains undetected.
In the UK, according to Campaign Against Living Miserably, around 1 in 4 people are now using AI tools for mental health advice, with the figure rising to 42% among Gen Z.
Index analysis
The study notes that AI chatbots designed to be helpful and agreeable inadvertently validate delusions rather than challenging them, turning a digital assistant into a potential catalyst for worsening psychiatric conditions.
Governance controls
The research team is calling for central regulation of AI technology similar to how social media is now being regulated to protect mental health.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.