AI Bias in Healthcare: What One Doctor’s Work Means for Patient Safety

Sonya Makhni, a doctor and researcher at the Mayo Clinic, is working to develop systems designed to reduce bias in medical artificial intelligence (AI) — the algorithms and automated tools increasingly used to support clinical decision-making. Her work focuses on health equity, meaning that AI-driven care should produce fair outcomes regardless of a patient’s background, ethnicity, or circumstances.

The concern is real. AI systems trained on historical medical data can reflect existing inequalities in healthcare, potentially producing recommendations that work less well for some patient groups than others. Makhni’s work aims to identify and reduce those risks before they affect care.

A central part of the effort is making sure clinicians can actually understand what the tools are telling them.

According to the Mayo Clinic, the goal is to give doctors and nurses AI outputs they can interrogate and trust — not black-box results they’re expected to accept without explanation. That transparency is seen as essential if AI is to be used responsibly in settings where decisions directly affect patient health.

The Mayo Clinic shared details of the work on its official social media channels, framing responsible AI development as a way to shape the future of patient care rather than simply automate it.

AI in medicine is an active area of research across the NHS and academic institutions worldwide. The NHS has published guidance on the safe deployment of AI tools in clinical settings, and the UK government’s AI Safety Institute has a broader remit covering high-risk uses of the technology — including in health.

—

If you have concerns about your health or care, contact your GP or call NHS 111. In an emergency, call 999.