Mount Sinai Study Reveals AI Bias Based on Patient Backgrounds

The study from the Icahn School of Medicine at Mount Sinai highlights significant disparities in AI-generated medical recommendations influenced by patients' socioeconomic and demographic backgrounds. Researchers tested nine large language models with 1,000 emergency cases, revealing that identical clinical situations sometimes led to different treatment suggestions based on factors like income, raising concerns about equity in AI healthcare applications. The findings stress the need for a framework to ensure AI tools are fair and reliable, aiming to safeguard against biased outcomes in medical care.

Opens on Mount Sinai