Key facts
- 77% of psychologists surveyed reported patients discussing AI for mental health support, diagnosis, or companionship.
- 39% of psychologists noted patients using AI for self-diagnosis.
- 35% of psychologists reported patients using AI as an additional mental health professional.
- 36% of psychologists observed patients developing dependency on chatbots.
- 15% of psychologists reported patients developing distorted thinking or delusions related to chatbots.
- AI models like xAI's Grok have been found to potentially reinforce delusions and suicidal ideation.
A survey by the American Psychological Association indicates a significant trend of patients incorporating AI chatbots into their mental health care. Of over 1,200 U.S. psychologists surveyed, 77% reported that their patients have discussed using AI for purposes such as emotional support, diagnosis, or companionship.
The survey revealed that 39% of psychologists observed patients using AI for self-diagnosis, while 33% noted patients using chatbots to assist with therapy or treatment. Furthermore, 35% of psychologists reported patients treating AI as an additional mental health professional.
While many patients find support through AI interactions, psychologists expressed considerable concerns. More than a third (36%) reported patients developing a dependency on chatbots, and 15% noted instances of distorted thinking or delusions linked to AI interactions. Psychologists also reported patients using AI for social purposes, with 22% using it for friendship and 13% engaging in intimate relationships with chatbots.
These findings align with recent research suggesting that several leading AI models, including xAI's Grok, may reinforce delusions, paranoia, and suicidal ideation. Psychologists largely share these concerns, with 97% believing chatbots could inadvertently reinforce negative behaviors or delusional beliefs, and 94% stating that current chatbots lack the necessary nuance for treating conditions.
The increasing use of AI in mental health comes as AI developers face growing legal scrutiny. Companies like OpenAI, Google, and xAI have been named in lawsuits alleging harm caused by their AI technologies. The APA emphasized that while AI can supplement professional care, it is not private and should not replace licensed mental health providers.