Trust Evolution in Agent and Multi-agent Systems: A Computational Modeling Perspective

The concept of trust is of fundamental importance in the field of multi-agent systems. Indeed, in this context, due to the intrinsic characteristics of the agent concept (autonomy within an environment with limited knowledge and control), trust has been introduced from the beginning as a basic theoretical requirement. In recent years, the application of this concept has exponentially increased, also thanks to the relatively recent emergence of trustworthy AI. Although, at this specific historical moment, this topic is capturing the attention of the scientific community with particular interest, trust has always been a key concept for social and non-social, human or artificial interaction. An extensive body of literature has developed around the concept of trust. Focusing on the world of multi-agent systems, we can observe that as the complexity of interaction grows, so does the importance of trust and the need to investigate its different types and faces. In this contribution, we aim to trace, through the analysis of some exemplary contributions presented in different editions of the Workshop WOA, the usage and evolution of the trust concept within multi-agent systems. The purpose of this analysis is twofold: on one hand, to understand how and why the choice to utilize trust was made, and on the other hand, to comprehend the future directions to be pursued and the challenges that lie ahead.