There has been a significant development that may impact healthcare providers using artificial intelligence (AI) in their practice. The Federation of State Medical Boards (FSMB) has recently released a report providing essential guidance to state regulators on the responsibilities of physicians when incorporating AI into patient care. The full report may be accessed here.
5 Key Take-Aways from the FSMB's New Guidance:
Why It Matters & What's Next
As AI continues to integrate into healthcare settings, understanding the legal and ethical implications becomes increasingly vital. The American Medical Association and other stakeholders are actively discussing the balance between embracing AI advancements and protecting both patients and healthcare providers from potential pitfalls associated with AI. Potential pitfalls include bias and inaccuracy, over-reliance on technology, and privacy concerns.
States are beginning to evaluate the role of medical boards in regulating AI use within medical practice, with some states starting to enact comprehensive AI statutes that directly and indirectly regulate the use of AI in health care. For example, in May of this year Colorado enacted a comprehensive state AI law that will go into effect on February 1, 2026, and which places certain limits and obligations on developers and entities that deploy AI including for use in health care services.1 Similarly, the Utah Department of Commerce's Office of Artificial Intelligence Policy has proposed regulations governing the use of AI chatbots and interactions by mental health care providers that is likely to be taken up by the Utah legislature later this year. Decisions and further regulatory frameworks are expected in the upcoming months, which could set precedents for how AI is integrated into healthcare nationally.
Speaking of AI . . . Use of AI in Recording Privileged and Confidential Peer Review Meetings:
While not discussed in the FSMB's report, considering the growing integration of AI tools in healthcare settings, it is important to address AI use in sensitive areas such as privileged and confidential peer review meetings. For example, if a practitioner relies on AI as opposed to exercising their professional medical judgment and this action jeopardizes patient care, peer review should be triggered. Further, using AI for recording and transcribing peer review meetings poses specific challenges and risks beyond a typical business meeting. Without carefully crafted protocols, there is a risk of breaching the confidentiality that is foundational to the peer review process, as AI generated meeting minutes or summaries may not be considered privileged material. Such breaches could undermine the legal protections typically afforded to these activities and potentially expose healthcare providers to legal liabilities. It is vital for healthcare institutions to establish and enforce strict guidelines that govern the use of AI in these sensitive settings, if AI is to be allowed at all, to maintain the integrity and confidentiality of the discussions.
As hospitals consider and develop AI policies and protocols we recommend they consider the following:
- Announce at the beginning of any meeting in which AI is being used to record and summarize the meeting, that AI will be used to do so, and that no other recordings of the meeting are authorized
- Require that any AI assisted documents will be marked accordingly and require review prior to being adopted as an official medical staff record
- Prohibit AI from being used to make decisions, and require that any decisions remain in the sole discretion of the medical staff based on independent verified information
- Require that all recordings are erased as a matter of routine within 30 days after the meeting
- If AI may be used to draft letters and memos, then require careful personal, (i.e., human) review of the AI generated document to ensure accuracy before being finalized. Any AI generated or assisted draft letter or draft memo must be marked accordingly
Footnote
1. Concerning Consumer Protections in Interactions With Artificial Intelligence Systems, SB 24-205, Colorado (enacted May 17, 2024).
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.