Studies often include both men and women, but they may not explore differences between the sexes. This gap can become especially important when old data is used to train artificial intelligence in healthcare. For example, many people learn CPR using manikins that look like men. A 2024 study of 20 CPR manikin models sold worldwide found that three-quarters were labeled as male or had no sex specified. Only one of the 20 models included a breast overlay, highlighting a broader tendency in medical training and research to use the male body as a standard. The lack of female representation in medical training and research can have real-world consequences. In a U.S. study of 19,331 out-of-hospital cardiac arrests, 39% of women who collapsed in public received bystander CPR, compared to 45% of men. Even a short delay in CPR can reduce survival chances. Another study found that people who received CPR four to five minutes after a cardiac arrest had a 27% lower chance of surviving to hospital discharge than those who received it within one minute. In England, fewer than one in 12 patients who received resuscitation from ambulance services survived 30 days, according to 2024 data. This lack of representation can also affect how AI systems are trained. Artificial intelligence in healthcare learns from medical records and research data, which may reflect past biases. If women and men with similar symptoms have historically been treated differently, an AI might learn and replicate those differences without understanding their origins. For instance, a study of medical students found that women with heart disease symptoms were less likely to be diagnosed or referred to cardiology than men, and their symptoms were more likely to be seen as psychological. An AI trained on such data might mistakenly treat this pattern as a reflection of actual disease. Efforts to improve representation in medical research have increased, but challenges remain. The 2016 Sex and Gender Equity in Research (SAGER) guidelines encourage researchers to consider sex and gender in their studies, but even when both sexes are included, results are often not analyzed separately. A 2026 review of 574 studies found that while 61% included both men and women, only 44% of those analyzed results by sex. Biological sex and gender can both influence health, but separating their effects is complex. Many health databases still use a single binary category for sex, limiting the understanding of how different factors affect health outcomes. As research continues to evolve, ensuring that both sex and gender are considered in medical training, research, and AI development is crucial for more accurate and equitable healthcare.