Research shows new reasoning language models o3-mini and DeepSeek-R1 still reproduce racial and gender stereotypes in clinical vignettes.
Key Details
- 1Flinders University researchers evaluated o3-mini and DeepSeek-R1 LLMs for bias in generating 36,000 clinical vignettes.
- 2Both models mirrored or exceeded the misrepresentation rates found in older models (up to 89% for race and 67% for gender).
- 3Overrepresentation of Black populations persisted for conditions like sarcoidosis and lupus—44% and 31% median misrepresentation, higher than GPT-4's 15%.
- 4Qualitative review showed models invoked demographic-disease associations without epidemiological justification.
- 5Findings published 28-May-2026, Journal of Medical Internet Research.
Why It Matters
Radiology and other specialties increasingly adopt LLMs for report drafting and workflow support. Persistent demographic biases in these systems risk perpetuating health disparities, underscoring the need for active bias monitoring and mitigation as AI tools integrate into clinical practice.
Source
EurekAlert
Related News
•EurekAlert
AI Model Combines ECG and Blood Tests to Spot Heart Transplant Rejection
An AI model integrating ECG data and blood biomarkers predicts heart transplant rejection with high specificity, reducing the need for invasive biopsies.
•EurekAlert
NIH Grant Funds AI Tool for Lung Cancer Discovery at Weill Cornell
Dr. Olivier Elemento receives an NIH Pioneer Award to develop an AI-human hybrid platform for cancer research, initially focusing on lung cancer using imaging and tissue data.
•EurekAlert
Study Shows AI Vision Misses Human-Like Perceptual Illusions
A York University study finds that AI vision systems lack the human-like perceptual shifts seen in biological vision during visual illusions.