Automated Repression

Tuesday, October 11, 2022
12:00 PM - 1:15 PM
(Pacific)

Virtual to Public. Only those with an active Stanford ID with access to E008 in Encina Hall may attend in person.

Eddie Yang seminar

This seminar will present the first evidence of systematic ethnic discrimination in Artificial Intelligence (AI) used to assist judges in criminal sentencing. Using audit experiments on a commercial criminal sentencing software, we find that AI predicts longer sentences for defendants with ethnic minority status and names that convey minority cues. The magnitude of discrimination is similar to existing findings from sentencing decisions of judges. Additionally, we find that AI may introduce new forms of discrimination not seen in human judgment.

ABOUT THE SPEAKER

Image
Eddie Yang
Eddie Yang is a PhD candidate in the Department of Political Science at UC San Diego and a predoctoral fellow at CDDRL and the Stanford Institute for Human-Centered Artificial Intelligence. His research focuses on repression and the politics of Artificial Intelligence. His dissertation studies how existing repressive institutions limit the usefulness of AI for authoritarian control, with a focus on China. His work has been published in both computer science and political science. 

Virtual to Public. Only those with an active Stanford ID with access to E008 in Encina Hall may attend in person.