Electronic Thesis and Dissertation Repository

Thesis Format

Alternative Format

Degree

Master of Health Information Science

Program

Health Information Science

Supervisor

Stark, Luke

Abstract

Artificial Intelligence (AI) has made remarkable strides in healthcare, from revolutionizing medical imaging, patient support, research insight, and health education. However, a recent shift from query-based to interactive and regenerative AI presents a question of properly facilitating its development. Upon close inspection, regulations and policy adaptations lag behind the rapidly advancing field of medical AI which raises concerns over governance and societal implications. This paper explains some of the technical and ethical considerations surrounding the use of health AI applications with chatbots as a specific example. It evaluates the existing policies and regulations in Ontario, Canada, that govern the production and use of chatbots in healthcare. I provide recommendations revolving around new policies in areas lacking the necessary framework with a goal not to limit the use of chatbots, but instead facilitate the adaptation of chatbots and AI in healthcare to reduce the potential harm resulting from improper and outdated policies.

Summary for Lay Audience

Artificial Intelligence (AI) has allowed machines to simulate human intelligence and are revolutionizing areas in healthcare rapidly. In healthcare, AI has been used to enhance processes related to problem-solving and decision-making, assisting in public health, medical research, and patient support. The recent development of an AI application called ChatGPT has raised the popularity of chatbots and its concept proved AI can successfully mimic human conversation. Moreover, it can provide recommendations by searching through an extensive library of resources and generating new content upon request. Chatbots have been used in healthcare as well including areas of mental health, patient support, and health education. However, there are ethical and safety concerns related to AI tools such as chatbots that are overshadowed by their benefits. Algorithmic biases and the “black box” have raised societal issues when AI tools are implemented. Governmental regulations also struggle to keep up with the rapid development of AI. This paper explains some of the technical and ethical considerations surrounding the use of health AI applications with chatbots as a specific example. It evaluates the existing policies and regulations in Ontario, Canada, that govern the production and use of chatbots in healthcare. It contains recommendations involving new policies in areas lacking the necessary framework. The intention of the policy analysis is not to limit the use of health chatbots, but instead facilitate the adaptation of chatbots and AI in healthcare to mitigate potential risks arising from outdated or inadequate policies.

Download

Share

COinS