Experts issue warning as more patients turn to AI chatbots for 'harmful' mental health advice

Psychologists have sounded the alarm over AI chatbots providing mental health advice to people, dubbing the information “unqualified and harmful”.

A survey conducted by the British Psychological Society (BPS) found that psychologists have noted a 49% increase in the number of patients seeking support from chatbots in the past year.

The poll, involving 197 participants comprised of psychologists, counsellors and psychotherapists, found that seven in 10 are somewhat or very concerned about the use of AI for mental health advice.

Nearly 88% (nine in 10) sounded the alarm that patients are relying too heavily on AI, while around 76% of survey participants said the technology is negatively affecting the mental wellbeing of their clients.

Some 80% told the survey they are concerned clients are becoming overly reliant on the technology, while around three quarters (76%) said it is negatively affecting client wellbeing.

Dr Laura Dean, president of the BPS, said: “This research shows there is real cause for concern for those that use AI for mental health therapy, and that concern is coming directly from psychological practitioners on the frontline providing vital in-person support.

“The risk of unregulated open access AI chatbots providing unqualified and harmful advice remains very real.”

Dr Dean warned that the integration of AI into healthcare “can only work with the right safeguards that protect both psychological professionals and the people they support”.

She added: “Although AI offers many benefits, it cannot replace the vital human element in mental health care.

“More research and insight from psychological practitioners, and people seeking support using AI tools, is needed to understand how and where human-led therapeutic services and AI can be safely integrated to deliver the right care and outcomes.”

One practitioner surveyed told the BPS: “I am concerned about not understanding exactly what AI does, how it works, and how it comes up with its answers.

“I am also concerned about people turning to AI because it feels “safer” (avoiding the relational risk of building trust with another human being), and we therefore increase the isolation and lack of social skills and contact in society, which is directly linked to quality of life and wellbeing.”

Another called for “accountability”, adding: AI often is seen as “free” but because of this, like with many social media platforms, the companies that run them have free rein to do as they please with little to no legal responsibilities.”

The BPS poll comes after a survey of 2,000 children aged 11-17 by Barnardo’s and the National Counselling and Psychotherapy Society (NCPS) found more than four in 10 (43%) had used AI for advice or emotional support.

Rosie Weatherley, information content manager at the charity Mind, said: “These survey findings echo what we are hearing from people and seeing through emerging evidence – that even casual use of generative AI can cause harm. So much is unknown about this new and ever-changing technology, particularly when it comes to the impact of long-term use.

“We’re supportive of digital tools that help people navigate information to make more sense of mental health and the services they can access, but these must not become a replacement for great quality care, where people feel properly heard and seen.

“We know that people still value human support over AI tools and we must not lose sight of this.”