A new American Psychological Association survey of over 1,200 U.S. psychologists reveals that 77% have patients who discussed using AI for emotional support, diagnosis, or companionship. Specifically, 39% reported patients using AI to self-diagnose, 35% used it as an additional mental health professional, and 33% used chatbots for therapy assistance. The survey highlights concerns about safety, privacy, and dependency: 36% noticed patient dependency on chatbots, and 15% observed delusions or distorted thinking linked to AI. While some patients reported positive experiences like feeling supported (68%) and improved coping skills (41%), the majority of psychologists (97%) worry chatbots may reinforce negative behaviors. The findings come amid legal scrutiny of AI companies, including lawsuits against OpenAI, Google, and xAI over alleged roles in real-world harm such as suicide and violent acts. The APA advises that AI is not a safe replacement for licensed professionals, especially for vulnerable populations.

Key facts

• 77% of psychologists have patients discussing AI for mental health.
• 39% of patients used AI for self-diagnosis; 35% as an extra professional.
• 36% of psychologists noticed patient dependency on chatbots.
• 15% observed delusions or distorted thinking linked to chatbots.
• 97% of psychologists fear AI reinforces negative behaviors.

KeyAudit data perspective