Cybersecurity News
Filters
Filtered by tag: bias × Clear
Report: AI reasoning gains fail to eliminate racial and gender bias in medical research
Matched: health
Researchers at Flinders University found that advanced AI reasoning models still reproduce racial and gender stereotypes when generating fictional medical patient cases, despite improvements in their analytical capabilities. The findings raise concerns about deploying such models in healthcare settings, where biased outputs could influence clinical decisions and reinforce existing disparities in medical treatment.