Mitigating Racial Bias in AI Research

Join us for the next workshop in the “Learning Together: Applying Equity, Diversity, and Inclusion Concepts in Research Ethics Review,” a monthly series bringing together experts from across the tri-campus on a range of research ethics topics from equity, diversity, and inclusion through a TCPS2 lens. This workshop will explore biases in both text and image understanding algorithms which have been criticized for being biased against racialized groups and others. 

 In this session, attendees will learn about:

• the difference between bias in data and bias in algorithms

• factors that lead to bias in research studies involving AI/ML

• the technical and ethical aspects of bias in image understanding algorithms including a brief survey of the sources of bias in image understanding algorithms

• considerations for weighing the potential risks and benefits of research when it is not possible to fully mitigate bias.