PSB 2026:

Fairness and Bias in Biomedical AI/ML: Defining Goals and Putting Them Into Practice

Overview

---

This interdisciplinary session will solicit research papers from disciplines such as ethics, medical humanities, social sciences, data science and biomedicine that develop understanding of different approaches and goals for how fairness is conceptualized and put into practice in medical AI/ML, as well as the practical challenges and trade-offs research teams face in addressing fairness, and facilitators of achieving fairness goals.

There is growing recognition that, despite fairness being a priority concern for the field of biomedical AI/ML, the variety in how fairness goals have been defined and addressed for projects can undermine those goals and needs examination in order to determine appropriate approaches. An interdisciplinary approach that includes ethics and social sciences will be valuable for examining different approaches to operationalizing fairness in projects for AI in healthcare. For example, quantitative fields tend to prioritize technical solutions to adjust algorithms to account for bias, while social scientists tend to view fairness through a lens of hierarchy and relational power dynamics, and bioethicists see fairness as based on rights. Thus, fairness may be conceptualized differently by different actors involved in developing AI/ML for biomedical research and healthcare. These different conceptions also interact with the different practices used to try to achieve fairness. These disconnects between different ways of conceptualizing, justifying, and operationalizing fairness, even within the development of the same project, contribute to the difficulties involved in putting fairness into practice in AI for healthcare and biomedical research.

A goal of this session is to facilitate an interdisciplinary conversation that examines practices in AI/ML that support fairness and mitigate bias and discusses best practices in developing AI/ML for healthcare to provide benefit across diverse populations. 

Session topics

---

We are interested in research from a range of disciplinary perspectives, including ethics and social sciences, that examines different approaches and goals for how fairness is conceptualized and put into practice in medical AI/ML, the practical challenges and trade-offs in fairness faced by research teams, and facilitators of achieving fairness goals.

Session organizers

---

Nicole Martinez-Martin, JD, PhD, Assistant Professor, Stanford Center for Biomedical Ethics

Mildred Cho, PhD, Professor, Associate Director, Stanford Center for Biomedical Ethics

*Primary contact:
Nicole Martinez-Martin, nicolemz at Stanford.edu

Submission Information

---

The submitted papers are fully reviewed and accepted on a competitive basis.