Chicago Booth’s Center for Applied Artificial Intelligence Receives a $21 Million Gift

A paper summarizing CAAI research recently received a Responsible Business Education Award from the Financial Times for academic research with impact. That research discovered large-scale racial bias in the algorithms used by insurers to predict the health-care needs of millions of US patients. CAAI researchers analyzed how patients were identified for a program offering extra medical support to people at high risk. Assuming past health-care costs indicated need, the algorithm overlooked the fact that Black patients cost the system less because they lacked access to or received lower levels of care, unfairly reducing by more than half the number identified for additional help. Based on this research, CAAI researchers developed a free online Algorithmic Bias Playbook for senior corporate executives, policymakers, and technical specialists in health care. The playbook has been downloaded and shared hundreds of times.

CAAI researchers have also studied the criminal justice area, finding that judges’ decisions about whether to release or detain an individual pretrial are affected by the faces they see in mug shots. Beyond the obvious features, like gender or skin color, the algorithm found that there are some additional features humans are predisposed to view favorably—like a clean-shaven or a rounder face—and that these predispositions can influence a judge’s decision. CAAI researchers were eventually able to predict, with 70 percent accuracy, what a judge’s decision would be—half of the model’s predictive power was driven by the mug shot alone.

While CAAI has already achieved much success, Emily Joy Bembeneck, the center’s director, believes the gift sets CAAI up for even greater future impact. One of the initiatives that could be made possible through this gift is the enhancement of the center’s resources beyond the classroom, such as the possibility of offering a library with additional assets to serve a much wider community.

“Ten years ago, people were not thinking about artificial intelligence and machine learning as they are now,“ Bembeneck said. “But it impacts every avenue of our lives. Algorithms use information about where we live, where we work, who we talk to, and what we like to guide decisions. From something as major as whether you get a mortgage or are admitted to a health-care program to something perhaps as trivial as what you see on social media, algorithms are there behind the scenes. Our role at the center is to understand how algorithms and humans together can make better decisions, and to help business and policy leaders apply those insights in ways that help make the world a better place.”