Understanding Public Judgements on Artificial Intelligence in Healthcare: Dialogue Group Findings From Australia

Introduction There is a rapidly increasing number of applications of healthcare artificial intelligence (HCAI). Alongside this, a new field of research is investigating public support for HCAI. We conducted a study to identify the conditions on Australians' support for HCAI, with an emphasis on identifying the instances where using AI in healthcare systems was seen as acceptable or unacceptable. Methods We conducted eight dialogue groups with 47 Australians, aiming for diversity in age, gender, working status, and experience with information and communication technologies. The moderators encouraged participants to discuss the reasons and conditions for their support for AI in health care. Results Most participants were conditionally supportive of HCAI. The participants felt strongly that AI should be developed, implemented and controlled with patient interests in mind. They supported HCAI principally as an informational tool and hoped that it would empower people by enabling greater access to personalised information about their health. They were opposed to HCAI as a decision-making tool or as a replacement for physician-patient interaction. Conclusion Our findings indicate that Australians support HCAI as a tool that enhances rather than replaces human decision-making in health care. Australians value HCAI as an epistemic tool that can expand access to personalised health information but remain cautious about its use in clinical decision-making. Developers of HCAI tools should consider Australians' preferences for AI tools that provide epistemic resources, and their aversion to tools which make decisions autonomously, or replace interactions with their physicians.