Navigating AI and Physician Advocacy in the Medical Space

Sponsored by ProAssurance

Artificial intelligence is no longer a futuristic concept but a rapidly evolving reality within modern medicine. As these technologies integrate into clinical practice, they offer opportunities to enhance care while simultaneously raising questions around patient safety, accountability, and the necessity of human oversight.

The Drive for Operational Efficiency

One of the primary advantages of AI in the current healthcare landscape is its ability to increase operational efficiency. “The most prevalent use case for AI that we’re seeing right now…across the board, across the country is ambient listening technology,” highlighted Bradley E. Byrne, Jr., J.D., CPHRM Southeast Regional Risk Manager at ProAssurance. This technology uses microphones or mobile apps to record patient encounters and automatically generate suggested clinical notes, billing codes, among other information. By synthesizing these interactions directly into electronic medical records (EMR), AI acts as a time-saving device that allows clinicians to remain more connected and focused on the patient during visits. This reduction in documentation time is a vital tool in combating physician burnout. “We’re [also] seeing [AI rollout] in hospitals when it comes to scheduling and creating workflows and trying to predict no-shows [or] ER peak times,” said Byrne.

AI’s impact is particularly visible in radiology, where it serves as a secondary reader to augment human accuracy. In dermatology, AI products are increasingly used to detect skin cancers and abnormalities. Beyond diagnostics, the FDA has recently approved AI systems for continuous post-surgical monitoring, providing an extra layer of safety when clinical staff are occupied. The implementation of AI is designed to not replace humans, but rather step in when human needs arise such as sleep.

Knowing the Risks

Despite these benefits, the transition toward clinical decision support tools introduces significant risks, such as the black box problem. This occurs when AI systems become so sophisticated that even their creators cannot fully explain how the software arrived at a specific conclusion. This lack of transparency poses challenges for informed consent, as physicians may struggle to explain the risks and alternatives of AI-driven recommendations to patients.

Furthermore, Byrne warns against automation bias, where providers may trust AI-generated data implicitly without performing due diligence. Much like a driver following a GPS without paying attention, a physician who fails to review AI-generated notes risks entering inaccuracies into a patient’s permanent record. Inaccurate documentation not only endangers patient safety but undermines a physician’s legal defense in medical malpractice cases, where the medical record serves as the primary evidence of care. “You want to be able to turn to the documentation and say this is what happened,” said Byrne.

A Call for Physician-Led Advocacy

To ensure these tools augment rather than supersede clinical judgment, physician advocacy is more important than ever. “One of the primary focuses of advocacy in the AI space…is to ensure AI doesn’t replace the physician’s medical expertise,” explained Lesley Lopez Viner, MS, CPHRM, Senior Risk Management Consultant at ProAssurance. Clinicians must have a voice in vetting new technologies and advocating for policies that prevent AI from being used by health plans to deny medically necessary care. Advocacy also plays a role in protecting physician identities from AI-generated deepfakes and safeguarding medical liability protections, so doctors are not held responsible for design flaws in AI systems.

Physicians can influence these regulations by joining state and local medical societies. “For example, I'm in Austin, Texas, and the Texas Medical Association here hosts a physician white coat invasion of the Texas Capitol every first Tuesday during legislative sessions. So, physicians, residents, and medical students gather at the Capitol in their white coats to visit with lawmakers about how proposed legislation can potentially affect the practice of medicine,” explained Viner. This sense of unity is extremely impactful. By showing up as a medical expert, you can explain to legislators the need to see beyond the data and statistics. This will help them understand how AI policy directly affects patient lives. “…Misinformation can spread faster than truth. Physician voices in legislator education and policy development when it comes to AI are critical,” said Viner.

Ultimately, AI in medicine is only as effective as the human oversight that governs it. By staying informed, engaging in advocacy, and calling for transparency from both developers and policymakers, physicians can use these tools to serve as a complement to clinical expertise rather than a substitute for it.