By Scott A. Hagen, JD and Brooke H. Davies, MPA
Originally published in Utah Physician Magazine, August/September 2024
Generative Artificial Intelligence in the Healthcare Setting
The positive impact of generative artificial intelligence1 (“AI”) on healthcare is just beginning. Generative AI increasingly supports healthcare providers with various tasks ranging from scheduling appointments to assessing a medical crisis and everything in between. In Utah, physicians are particularly benefitting from generative AI’s time-saving role in clinical noting and documentation.2
Despite its utility and benefits, as AI’s role in healthcare expands and improves, medical providers should stay apprised of the evolving legal disclosure requirements for using generative AI in healthcare to avoid costly fines or litigation. Effective May 1, 2024, Utah’s newly passed “Generative Artificial Intelligence” law requires disclosure of generative AI use in regulated professions like healthcare.3 Utah Code section 13-2-12(4)(a) states that “[a] person who provides the services of a regulated occupation shall prominently disclose when a person is interacting with a generative artificial intelligence in the provision of regulated services.” In other words, when a healthcare provider uses generative AI in their practice, upfront and clear disclosure to the patient prior to that use is necessary.
What Disclosure is Required to Patients When Healthcare Providers Use Generative AI?
The compulsory disclosure of generative AI’s use in regulated industries like healthcare is more proactive than disclosure in non-regulated industries. In a regulated industry, disclosure of generative AI use must be provided “verbally at the start of an oral exchange or conversation” and “through electronic messaging before a written exchange.”4 Importantly, this means that for a telehealth or in-person visit, a healthcare provider may not simply include disclosure of generative AI use with other privacy notices– someone must verbally inform the patient of AI use prior to the provision of care. Along the same lines, in a written exchange between a provider and patient, such as through a provider-patient messaging system, a message disclosing AI’s use must come before healthcare communication between the
parties commences.
As an example of the unique disclosure the new law requires, if a physician employs an AI program like DAX Copilot during a patient visit to “capture[] exam room or telehealth conversations” that then “creates draft clinical summaries,” the patient may not simply read through and submit an AI disclosure form while in the waiting room.5 Instead, the statute seems to require that the provider verbally inform the patient, perhaps just before the patient enters an exam room, but in advance of any use of DAX Copilot, that the patient will be “interacting with generative AI” during the visit.6 The nuanced details of what the disclosure law specifically requires are currently unclear. Questions remain surrounding exactly what AI programs the law encompasses,7 who needs to disclose AI’s use during a visit (e.g., a medical assistant, physician, or both), how often disclosure is required, and whether disclosure is needed if using AI only for post-visit activities like notation summaries or referral letters. In part to address the rapidly
changing AI landscape, the Utah legislature created America’s first “Office of Artificial Intelligence
Policy”8 alongside the new disclosure law, and this office will play an important role in future clarification and support.9
Current Best Practice Recommendation
Because specific terms under the new law are not yet clear (e.g., What does it mean for AI to “interact with a person”? Is using AI for post-visit documentation considered a “provision of regulated services”?), for now, engaging in upfront disclosure with patients when providers may use AI at any stage of the healthcare visit is recommended. By beginning with a verbal AI disclosure during in-person and telehealth visits or an upfront message when communicating through written exchanges, providers will better protect themselves from fines and litigation while still benefitting from the positive uses of AI in healthcare. As this article does not address all the varied technologies, nuances, and lightning-speed developments of generative AI in healthcare, providers should consistently consult with trusted legal counsel to remain up-to-date and compliant with the evolving legal structures surrounding generative AI.
Endnotes
1. Bernard Marr, The Difference Between Generative AI and Traditional AI: An Easy Explanation for Anyone, www.forbes.com/sites/bernardmarr/2023/07/24/the-difference-between-generative-ai-and-traditional-ai-an-easy-explanation-for-anyone
2. Giles Bruce, How University of Utah Health Physicians Fell in Love with AI, Becker’s Health IT www.beckershospitalreview.com/innovation/how-university-of-utah-health-physicians-fell-in-love-with-ai.html
3. Utah Code § 13-2-12
4. Utah Code § 13-2-12(5)
5. Introducing DAX Copilot: Your AI-Powered Clinical Companion, www.voiceautomated.com/dax-copilot-dragon-ambient-experience
6. Example based on a conservative interpretation of the statute. Under a different interpretation, it is possible that the statute only requires this type of disclosure when the patient is solely interacting with generative AI and not another human being.
7. Practical Law Data Privacy & Cybersecurity, Utah AI Consumer Protection Compliance and AI Policy Act Checklist (on file with Thomson Reuters Westlaw, W-043-1595)
8. Utah Code § 13-72.
9. Martha Harris, Utah’s New Office of Artificial Intelligence Will First Focus on Mental Health Care and AI, www.kuer.org/politicsgovernment/2024-07-09/utahs-new-office-of-artificial-intelligence-will-first-focus-on-mental-health-care-and-ai
Scott A. Hagen
moc.nqr@negahs
801-323-3328
Scott A. Hagen, JD, is the Chair of the Firm’s Healthcare Law Section and the former Chair of the Employment and Labor Section. His practice includes labor relations, employment litigation, employee benefits (ERISA) litigation, and client representation before administrative agencies.
Brooke H. Davies
Brooke H. Davies, MBA and JD Candidate 2026, is a second-year law student at the University of Utah, S.J. Quinney College of Law, and RQN Summer Associate. A Utah native, Brooke hopes to spend her career as an attorney supporting those serving Utah communities.
