ABSTRACT:
Background
With the rapid evolution of artificial intelligence, the work of health care professionals is changing quickly. Both human-in-the-loop and human-out-of-the-loop artificial intelligence are becoming prevalent in patient care, and this raises questions regarding the role of health professionals' regulators in ensuring safe and effective practice and safeguarding the public's interests.
Objectives
This research explored regulators' perspectives on whether and how regulatory tools could (or should) be used with respect to artificial intelligence.
Methods
An artificial intelligence-enabled search of websites identified regulatory leaders with respect to regulation of artificial intelligence. Eighteen regulators from diverse health professions in the US, Canada, and the UK were interviewed, based on their documented work in this area.
Results
Key findings included: a) recognition that the rapid evolution of artificial intelligence was eclipsing regulators' abilities to manage it; b) belief that current regulatory tools were sufficient to manage risks of human-in-the-loop artificial intelligence, c) belief that regulation of human-out-of-the-loop artificial may be infeasible for regulators; and d) focus on development of educational, principles-based guidance to support practitioners in responsible adoption of artificial intelligence in their practice (rather than a rules-based approach).
Conclusions
As artificial intelligence continues to evolve rapidly and permeate professional and daily life, further research is necessary to ensure public interests are safeguarded effectively, in ways that do not stifle technological innovation and growth.
