Newer AI models still reproduce racial and gender stereotypes in medicine
Flinders University researchers have evaluated two next-generation reasoning large language models (LLMs)—o3-mini and DeepSeek-R1—and found that when asked to describe fictional patients with common medical conditions, the models frequently reproduced racial and gender stereotypes, indicating that advancements in AI reasoning do not inherently improve representational fairness.
Researchers at Flinders University have discovered that two cutting-edge AI models, o3-mini and DeepSeek-R1, continue to reproduce racial and gender biases in the portrayal of medical conditions. Despite advancements in AI reasoning, these models still overrepresented Black patients in stereotypical conditions like sarcoidosis and preeclampsia at rates comparable to earlier generations like GPT-4.
This finding suggests that improvements in AI reasoning do not necessarily translate to better representation fairness, highlighting the need for ongoing monitoring and mitigation of biases in large language models used in healthcare.
Written by urgent.news from Medical Xpress's reporting — not their text. Machine-written; read the original for the full account.
