Nearly half of the participants in a U.S. survey viewed AI as beneficial for mental healthcare, though concerns around incorrect diagnoses and reduced interaction with providers remain significant.
A recent study from Columbia University School of Nursing highlighted that, while AI adoption in healthcare is growing, limited research has explored patient perspectives, especially in mental healthcare. Previous studies mainly focused on somatic healthcare issues like perinatal health and radiology, with patient trust hinging on the use case and clinician endorsement.
The survey, which included 500 U.S. adults, revealed that 49.3% believed AI could be beneficial in mental healthcare, though opinions varied by demographic. Black respondents and those with lower health literacy were more likely to see the benefits, while women were less inclined to share that view.
Major concerns included AI’s accuracy, risk of incorrect diagnoses, potential for inappropriate treatments, and fear of losing personal connection with providers. Additionally, most participants (81.6%) believed that mental health misdiagnoses involving AI would remain the provider’s responsibility.
Key values identified by respondents included confidentiality, autonomy, and the ability to understand personal mental health risk factors. The researchers emphasized the need to communicate AI tool accuracy and ensure trust between patients and providers when implementing AI in mental healthcare.
Lead researcher Dr. Natalie Benda emphasized the importance of understanding patient perspectives, as AI becomes more prevalent, to ensure safe and effective deployment of AI tools in mental health.