In Trust We Survive: Emergent Trust Learning

Abstract

We introduce Emergent Trust Learning (ETL), a lightweight, trust-based control algorithm that can be plugged into existing AI agents. It enables these to reach cooperation in competitive game environments under shared resources. Each agent maintains a compact internal trust state, which modulates memory, exploration, and action selection. ETL requires only individual rewards and local observations and incurs negligible computational and communication overhead. We evaluate ETL in three environments: In a grid-based resource world, trust-based agents reduce conflicts and prevent long-term resource depletion while achieving competitive individual returns. In a hierarchical Tower environment with strong social dilemmas and randomised floor assignments, ETL sustains high survival rates and recovers cooperation even after extended phases of enforced greed. In the Iterated Prisoner's Dilemma, the algorithm generalises to a strategic meta-game, maintaining cooperation with reciprocal opponents while avoiding long-term exploitation by defectors. Code will be released upon publication.

Figures (11)

• Figure 1: Overview of the Tower environment and Emergent Trust Learning (ETL). A food platform starts at the top of a vertical tower and moves down through the floors, each occupied by one learning agent (randomly placed for every run). On each floor, the current agent may eat part or all of the remaining food, which directly improves its own short term reward but leaves less for agents on lower floors. If agents act too greedy, those on upper floors consume too much and agents below eventually starve. As there are no explicit rewards fostering agent cooperation, it has to emerge from behaviour. ETL equips each agent with an internal trust state, memory of recent outcomes, and an adaptive exploration rule based only on local information, thus agents gradually learn to moderate their consumption and maintain a more cooperative and survivable system.
• Figure 2: Schematic of the memory mechanism in ETL, highlighting how short-term and long-term experience buffers are updated from state-action-reward tuples and jointly drive subsequent action decisions.
• Figure 3: Overview of the trust mechanism in ETL: locally observed cooperation and dissatisfaction signals update a scalar trust state, which is smoothed over time and fed back into decision making without requiring explicit communication or identities.
• Figure 4: Exploration mechanism in ETL, where the exploration rate is adapted based on trust stability, perceived risk and recent performance, so that agents explore more under uncertainty and exploit more in stable, trusted regimes.
• Figure 5: ETL reduces long-run conflict intensity compared to a trust-free baseline, yielding substantially fewer low-value clashes over shared resources across 30 independent runs.