Artificial intelligence: danger or hope for our mental health?

Artificial intelligence: danger or hope for our mental health?

Artificial intelligence (AI) is increasingly influencing people's everyday lives and raising many questions, especially with regard to mental health. A recently held panel discussion at the Marsilius Kolleg at Heidelberg University addressed this topic and focused on the effects of AI on self-image and social relationships. Both the opportunities and challenges associated with digital technologies were discussed in detail. Particularly in the context of the COVID-19 pandemic, which has made access to healthcare services more difficult for many people, the use of AI is seen as a potential means to improve mental health care. According to a report by econstor.eu In 2021, there were over 150 million people living with mental illness in the WHO European Region.

The discussion about the burdens of digital surveillance and automated decision-making processes was a central topic of the event. The experts, including Michael Gertz, Julia Peters and Svenja Taubner, highlighted the different perspectives and presented the potential of AI in the diagnosis and treatment of mental illnesses. Examples of use include early recognition of symptoms and support of therapeutic processes.

Limits and potential of AI

The events as part of the “Marsilius controversial” series are aimed specifically at a broad audience in order to discuss socially relevant questions in an interdisciplinary manner. One goal is to invite people to actively participate in discussions about the impact of AI on mental health. The moderation was by the directors of the Marsilius Kolleg, Prof. Dr. Friederike Nüssel and Prof. Dr. Michael Boutros, took over and promoted the dialogue between different scientific cultures.

However, the challenges in using AI were also discussed in a new WHO study. This highlights deficits in the use of AI in mental health care and research. The study highlights methodological and qualitative shortcomings in the use of AI in research on mental disorders between 2016 and 2021, with a clear focus on depressive disorders and schizophrenia. In many cases, the use of AI remains unbalanced, and there are significant gaps in understanding how AI can also be used to research other mental health issues, as reported by who.int mentioned.

Another key finding of the WHO study is that AI applications are often complex and, if used improperly, can lead to distortions and inaccurate results. Methodological deficiencies such as insufficient data validation and lack of risk evaluation have been identified as inhibiting factors that could make practical implementation of AI difficult. There is a lack of transparent reporting on AI models, which affects reproducibility and collaboration between researchers.

The WHO/Europe Action Plan for 2023 to 2030 recognizes the need for innovation in predictive analytics through big data and AI. Given the challenges and opportunities that artificial intelligence offers in mental health, it remains important to continue to promote and deepen the dialogue between science and society.