From Intelligent Agents to Trustworthy Human-Centred Multiagent Systems

TL;DR

This paper reviews the University of Southampton's research on multiagent systems, emphasizing trustworthy autonomous and human-centered AI systems, highlighting key scientific contributions and future challenges.

Contribution

It summarizes the group's extensive work in learning, coordination, and formal methods for trustworthy and societal-beneficial multiagent systems.

Findings

Significant advances in multiagent learning and coordination

Development of formal methods for reasoning in MAS

Identification of open challenges in trustworthy AI deployment

Abstract

The Agents, Interaction and Complexity research group at the University of Southampton has a long track record of research in multiagent systems (MAS). We have made substantial scientific contributions across learning in MAS, game-theoretic techniques for coordinating agent systems, and formal methods for representation and reasoning. We highlight key results achieved by the group and elaborate on recent work and open research challenges in developing trustworthy autonomous systems and deploying human-centred AI systems that aim to support societal good.