The rapid advancement of artificial intelligence (AI)-based technologies during the last decade has led to dramatic innovation in healthcare technology across a wide variety of functional areas and clinical domains. Recent advances in generative AI have led to the development of modern large language models (LLMs), such as OpenAI’s ChatGPT. Numerous "chatbot" applications using LLM technology are available on the market, and some claim to provide therapeutic benefits for mental health. Chatbots currently in use include direct-to-consumer wellness products, tools used by health systems for screening and referral, and service options in employee assistance programs. However, the promise of technology using LLMs is predicated on the quality and reliability of the models’ generated outputs. The inherent uncertainty of these outputs generates ethical questions that psychiatrists will increasingly face as we and our patients interact with AI technologies.
The session will begin with an introduction to AI in healthcare and definitions of frequently used AI-related terminology. We will then present an overview of how AI is already being used in healthcare with a focus on behavioral health applications, including risk assessment, diagnosis and monitoring, documentation, and workflow tools. Next, the panel will explore the ethics of AI in behavioral health, focusing on the use of AI in psychotherapy. Historical background on the use of conversational agents for therapy will be presented before moving to some of the applications available in today’s market. The panel will present a case vignette to illustrate how a psychiatrist may have to wrestle with the ethics of AI when caring for a patient and considering whether to recommend chatbot therapy.
The panelists will discuss ethical issues related to AI in psychotherapy in terms of the four ethical principles: beneficence, nonmaleficence, autonomy, and justice. As part of this discussion, the panel will explore risks of bias and misinformation in AI-based algorithms. The panel will consider how the ethical considerations apply to the case vignette. An open discussion and Q&A with the audience will follow. Finally, the chair will outline some resources on AI for audience members who wish to learn more.
@inproceedings{he2024ethics_ai,
author = {He, C. X. and Sarma, K. V. and Hu, R.},
title = {The Ethics of Artificial Intelligence in Psychiatry: A Beginner’s Exploration},
booktitle = {APA Mental Health Services Conference},
year = {2024},
}