Imagine several self-driving cars attempting to navigate an intersection without stopping, a surgeon operating with the aid of a robotic nurse, or members of Parliament casting votes on a bill. In each of these cases, a collective decision is made by multiple agents. But who is responsible when something goes wrong? The answer is often unclear. Sometimes no one can be individually blamed; at other times, responsibility is diffused among many, each pointing to someone else in a circle of blame.
In my work, I aim to pioneer a new field—Responsible Mechanism Design—that bridges the sciences and the humanities by drawing on tools from game theory, ethics, logic, and law. I study how responsibility is allocated within collective decision-making systems and how it is shaped by social norms as well as by agents’ knowledge, beliefs, and trust. I analyse the responsibility-related properties of existing mechanisms and design new ones that promote fairness, traceability, and individual accountability. Much of my work is carried out in close collaboration with colleagues and students—from California to China—whose contributions have been instrumental in shaping this emerging area of research.
Selected Publications
P. Naumov, J. Tao, Responsibility Gap in Collective Decision Making, 34th International Joint Conference on Artificial Intelligence (IJCAI-25)
Q. Shi, P. Naumov, Responsibility in Multi-Step Decision Schemes, Journal of Philosophical Logic, 2025