Legal and Ethical Challenges in Integrating Artificial Intelligence into Clinical Practice: A Qualitative Study of Physicians’ Real-World Experiences
ABSTRACT
Background:
The adoption of artificial intelligence (AI) in healthcare has accelerated, yet physicians continue to face substantial legal, ethical, and regulatory uncertainties when considering AI integration into clinical practice. Although literature on AI in healthcare is expanding, there is limited insight into the real-world concerns voiced by clinicians navigating these uncharted territories.
Objective:
This study explores the legal and ethical uncertainties raised by Canadian physicians in relation to AI use in clinical care, using actual medico-legal advice requests as a window into their practical concerns.
Methods:
We conducted a comprehensive thematic analysis of 46 medico-legal advice calls made by physicians to a national medico-legal advisory service between March 2023 and February 2025. The calls were analyzed to identify key themes and patterns in physicians’ questions and perceived risks regarding AI tools in clinical contexts.
Results:
Eight key themes emerged, including the use of AI scribes, data privacy and security, patient consent, data ownership, regulatory uncertainty, medico-legal liability, vendor agreements, and concerns about accuracy and bias. Many of the inquiries focused on administrative and documentation-related AI applications rather than diagnostic tools, reflecting the current stage of AI integration in everyday clinical workflows. Physicians expressed uncertainty regarding legal responsibility, alignment with privacy laws, and appropriate communication with patients about AI use.
Conclusions:
This study offers unique insight into frontline physicians’ real-time concerns about AI, highlighting the need for clearer regulatory guidance, clinical standards, and legal frameworks to support safe and ethical AI adoption in healthcare.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.