Last August, the Notre Dame-IBM Technology Ethics Lab released its second annual Call for Proposals (CFP) for the purpose of funding practical and applied interdisciplinary research in tech ethics, ultimately selecting 19 projects from investigators representing nine countries for awards. Projects are being completed throughout 2023.
The theme for this CFP, Auditing AI, speaks to the need to rigorously examine artificial intelligence systems so as to take seriously the variety of concerns around them. These concerns can range from well-documented issues with current tools to future threats to the existence of humanity itself, even as the vast majority of people working in the field still consider human extinction as the direct result of AI to be in the realm of science fiction.
While award-winners from the first CFP had the chance to interact with each other through virtual sessions, this second cohort is the inaugural group to experience a new element of the program:
An on-campus workshop designed to allow recipients to connect in person, share progress on their projects, and learn from one another.
The workshop kicked off on Tuesday, June 13, with dinner and a keynote address from Marianna Bergamaschi Ganapini titled “Building an Infrastructure of Trust for Autonomous Systems.” An assistant professor of philosophy at Union College, Ganapini will begin a one-year term as a visiting fellow of the lab on September 1.
Day two started with a keynote on “Auditing AI in Practice” from Heather Domin, program director for AI Ethics and Tech Ethics by Design at IBM and the IBM associate director of the lab.
“The diversity of perspectives and issues highlighted by the group reflects the current complexity of auditing AI,” Domin said. “It is important that we remember why we audit AI and continue to press through the current challenges to help ensure trustworthy AI systems. The award recipients' work is also instrumental to understanding and identifying cutting-edge solutions and potential paths forward.”
Domin’s talk was followed by two morning sessions of presentations from CFP project principal investigators (PIs), an interactive keynote on “TRUST: Building a Sustainable Future for People and Machines” led by Michael Hemenway and Jeni Rinner of the Iliff School of Theology, and then two more rounds of presentations by project PIs in the afternoon.
After Wednesday night’s dinner, attendees engaged in an open information-sharing and brainstorming plenary session. Over the course of the workshop, there were also opportunities for informal networking and tours of the Notre Dame campus.
Among the PIs who presented were Sharanya Shanmugam, Willow Wong, and Zhang Wenxi of the Centre for AI and Data Governance at Singapore Management University (SMU). Together with their SMU colleague Mark Findlay, they are pursuing a project on “AI Audits for Whom? Asian Perspectives on Rebuilding Public Trust via Community Ethics and Conflict Resolution Mechanisms.”
“It was truly a joy to converse with researchers from such a diverse range of academic backgrounds, all seeking to ground the purpose of auditing AI within the unique contexts of their domains,” Wong said. “As an early career researcher, I felt very supported in sharing my thoughts and receiving generous feedback from the other participants during the workshop. I also appreciated the opportunity to have more informal conversations over a meal, which gave everyone a chance to think about the intricate relationship between technology and society outside of—quite literally—the seminar room.”
This CFP is part of the lab’s broader Auditing AI Initiative that also has led to the creation of an undergraduate course on the topic that will be offered for the first time this fall and can be counted toward Notre Dame’s undergraduate minor in tech ethics.
Project descriptions for all 19 projects are available at techethicslab.nd.edu/call-for-proposals. The lab’s next Call for Proposals and instructions for applying for funding will be published later this summer.