May 16 2019
In the future, interactions of humans with artificial intelligence (AI) will turn out to be a highly common attribute of their lives. Currently, researchers from the Technical University of Munich (TUM) have completed the first investigations of how “embodied AI” could be helpful in treating mental illness.
Professor Alena Buyx has looked into ethical questions raised by treating mental illness with AI. (Image credit: Andreas Heddergott/TUM)
They conclude that critical ethical questions related to this technology have still not been answered. There is an emergent need to act on the part of researchers, professional associations, and governments.
There already exist several initiatives using embodied AI to enhance mental health: computer-generated avatars helping patients to cope with hallucinations, robot dolls teaching autistic children to communicate better, and virtual chats that offer support with depression. These applications are called embodied since they entail interactions between an artificial agent and individuals, leading to entirely new dynamics.
The application of AI in psychotherapy is nothing new. In the 1960s, the first chatbots formed the illusion of a psychotherapy session. However, actually, this was little more than a gimmick. Using the existing higher computing power and advanced algorithms, considerably more is possible.
The algorithms behind these new applications have been trained with enormous data sets and can produce genuine therapeutic statements.
Alena Buyx, Professor of Ethics in Medicine and Health Technologies, TUM
Along with Dr Amelia Fiske and Peter Henningsen, Professor of Psychosomatic Medicine and Psychotherapy, Buyx performed the first systematic survey of embodied AI applications for mental health. She has also drawn conclusions on the associated challenges as well as opportunities.
Access to Treatment for More People
The new applications exhibit immense potential. They can render treatment available to much more people since they are not restricted to specific locations or times. Moreover, certain patients feel it is easier to interact with AI compared to interacting with a human being. However, there are associated risks as well. “AI methods cannot and must not be used as a cheaper substitute for treatment by human doctors,” stated Amelia Fiske.
Although embodied AI has arrived in the clinical world, there are still very few recommendations from medical associations on how to deal with this issue. Urgent action is needed, however, if the benefits of these technologies are to be exploited while avoiding disadvantages and ensuring that reasonable checks are in place. Young doctors should also be exposed to this topic while still at medical school.
Peter Henningsen, Dean, TUM School of Medicine
Ethical Rules for Artificial Intelligence Still Lacking
Currently, researchers have been making heightened efforts to come up with guidelines for AI, such as the Ethics Guidelines for Trustworthy AI recently published by the European Union. Yet, Buyx, Fiske, and Henningsen also notice an emergent need for regulation of AI application in specialized fields.
Therapeutic AI applications are medical products for which we need appropriate approval processes and ethical guidelines. For example, if the programs can recognize whether patients are having suicidal thoughts, then they must follow clear warning protocols, just like therapists do, in case of serious concerns.
Alena Buyx, Professor of Ethics in Medicine and Health Technologies, TUM
Moreover, a comprehensive study is required regarding the social impacts of embodied AI. “We have very little information on how we as human beings are affected by contact with therapeutic AI,” stated Alena Buyx. “For example, through contact with a robot, a child with a disorder on the autism spectrum might only learn how to interact better with robots—but not with people.”