Cooperative Resilience | Manuela Chacon-Chamorro

Sustaining Collective Welfare Under Disruption: A Pilot Study of Cooperative Resilience in Human and LLM Multi-Agent Systems

Thu, 10 Sep 2026 00:00:00 +0000

✨ About the talk

At FAIEMA 2026, I presented our work “Sustaining Collective Welfare Under Disruption: A Pilot Study of Cooperative Resilience in Human and LLM Multi-Agent Systems.” This research was developed jointly with Juan Sebastián Pinzón, Rubén Manrique, Luis Felipe Giraldo, and Nicanor Quijano at Universidad de los Andes.

The study explores how human groups and LLM-based multi-agent systems respond when cooperation is challenged by environmental and behavioral disruptions. Both populations interacted in the same mixed-motive social dilemma, following a shared experimental and evaluation protocol.

Presenting our work on cooperative resilience in human and LLM-based multi-agent systems.

🔎 Main findings

Our results showed that human groups sustained higher levels of cooperative resilience across the evaluated scenarios. Human participants tended to regulate their use of shared resources and maintain collective welfare more effectively under disruption.

In contrast, LLM-based agents showed more limited adaptation and tended toward higher resource consumption. Communication improved their performance in some scenarios, but it did not fully close the resilience gap between human and LLM-based groups.

These findings highlight the importance of evaluating AI agents not only under stable conditions, but also before, during, and after adverse events.

🌍 The conference experience

Participating in FAIEMA 2026 was a very rewarding experience. I had the opportunity to share our research with people from different institutions and research backgrounds, receive thoughtful questions, and exchange ideas about cooperative AI, multi-agent systems, resilience, and the societal implications of artificial intelligence.

Beyond the presentation itself, the conference created a welcoming environment for meeting other researchers and learning about work across different areas of AI. Sharing the conference activities and dinner with so many people made the experience especially enjoyable and memorable.

Sharing ideas and experiences with other participants during FAIEMA 2026.

💭 Final reflection

Presenting this work in L’Aquila was particularly meaningful because it brought together several dimensions of my doctoral research: defining cooperative resilience, studying collective behavior under disruption, and comparing the capabilities of humans and artificial agents.

The experience also reminded me that conferences are not only spaces for presenting results. They are opportunities to discuss unfinished questions, receive new perspectives, build connections, and recognize how research grows through interaction with others.

Sustaining Collective Welfare Under Disruption: A Pilot Study of Cooperative Resilience in Human and LLM Multi-Agent Systems

Thu, 10 Sep 2026 00:00:00 +0000

Version note. This entry corresponds to the revised camera-ready manuscript accepted at FAIEMA 2026. It is distinct from the earlier arXiv preprint, “Evaluating Cooperative Resilience in Multiagent Systems: A Comparison Between Humans and LLMs,” which remains available as the original preprint. The accepted manuscript is a substantially revised and polished version of the work and will be linked here once it becomes available in the official Springer proceedings.

Abstract. The ability to sustain collective welfare under disruption is key for deploying agents in real-world multi-agent environments. However, most evaluations focus on stable scenarios; as a result, agent behavior under adverse conditions remains poorly understood. In this work, we present a pilot comparative study of cooperative resilience—the ability of a system to act before, during, and after a disruption to preserve collective welfare—between human groups and LLM-based multi-agent systems operating in the same social dilemma under a shared disruption and evaluation protocol. Both populations are evaluated with and without communication, using population-specific observation and communication interfaces while preserving the same task structure, disruption scenarios, and evaluation criteria. Our results indicate that, in this setting, LLM-based agents show more limited adaptation and lower levels of collective welfare compared to human groups. These findings suggest potential limitations in current agent designs and motivate the development of more robust and adaptive multi-agent systems.

Learning Incentive Structures for Cooperative Resilience in Multi-Agent Systems under Social Dilemmas

Thu, 29 Jan 2026 00:00:00 +0000

Abstract. Multi-agent social dilemmas, such as the tragedy of the commons, capture settings where individual incentives conflict with collective well-being, making these systems highly vulnerable to collapse under disruptions. In this context, this work studies cooperative resilience, understood as the system-level ability to maintain collective well-being under perturbations through adaptive agent behavior. We propose a framework for learning incentive structures aligned with collective well-being in multi-agent reinforcement learning systems, where reward functions shape individual decision-making and collective behavior. A resilience metric is used to score and rank agent trajectories, allowing the inference of reward functions that promote resilient collective behavior. These inferred reward functions are integrated into the multi-agent reinforcement learning process to shape agent interactions in social dilemma settings. The approach is evaluated in resource-sharing environments subject to disruptions, using three incentive structures: individual incentives, resilience-aligned incentives, and a hybrid incentive structure that combines both individual and collective components. The results show that the hybrid incentive structure promotes sustained collective behavior, reduces collapse events associated with resource depletion, and preserves system performance under disruption. These findings highlight the role of incentive design as a mechanism for promoting resilient collective behavior and provide a computational framework for multi-agent social dilemmas under disruptions.