Navigation auf uzh.ch
Jonathan Fu, Annette Krauss, and Mrinal Mishra, in partnership with Women’s World Banking and several financial institutions in India, Colombia, and Mexico, conducted a two-year technical assistance project, with funding support from Data.org, the Mastercard Center for Inclusive Growth, and The Rockefeller Foundation. The researchers conducted comprehensive audits of the partner financial institutions’ credit portfolios or particular product lines for signs of algorithmic- or human-driven bias. They co-authored a practical field guide and helped develop an open-source Python-based toolkit that allows users to assess their credit underwriting processes for signs of such biases based on different concepts of “fairness”. The researchers are currently working on academic research papers to improve and disseminate knowledge on underlying sources of bias and potential ways to mitigate them.