Spotlight
in
Workshop: Multi-Agent Security: Security as Key to AI Safety
Defining and Mitigating Collusion in Multi-Agent Systems
Jack Foxabbott · Sam Deverett · Kaspar Senft · Samuel Dower · Lewis Hammond
Keywords: [ collusion ] [ mechanism design ] [ Multi-Agent Systems ] [ multi-agent reinforcement learning ]
Sat 16 Dec 7 a.m. PST — 3:30 p.m. PST
Collusion between learning agents is increasingly becoming a topic of concern with the advent of more powerful, complex multi-agent systems. In contrast to existing work in narrow settings, we present a general formalisation of collusion between learning agents in partially-observable stochastic games. We discuss methods for intervening on a game to mitigate collusion and provide theoretical as well as empirical results demonstrating the effectiveness of three such interventions.