This perspective clarifies where current alignment methods genuinely benefit from game-theoretic analysis, where the framework is looser, and what challenges remain in building robust, adaptive, and verifiable AI systems.
Abstract
As large language models and increasingly capable AI agents are deployed in high-risk settings, aligning them with complex human values has become a central challenge. Existing alignment methods, while effective in improving helpfulness, harmlessness, and controllability, often struggle to capture real-world preferences that are context-dependent, non-transitive, and shaped by dynamic multi-party interactions. This survey reviews AI alignment through a game-theoretic lens. Specifically, it organizes recent progress around key game-theoretic elements and synthesizes the literature along three challenges: preference diversity, alignment priority, and temporal dynamics. This perspective clarifies where current alignment methods genuinely benefit from game-theoretic analysis, where the framework is looser, and what challenges remain in building robust, adaptive, and verifiable AI systems.
It is proved that in closed, strictly competitive games enough computation drives the human to the classical security value and creativity is worthless; that creativity has value exactly when an unmodeled move beats the AI’s best defense.
Stochastic Reflective Memory Ascent (SRMA), which accepts a candidate memory only after a grounded evaluation risk strictly decreases, is introduced and provides confidence gating for stochastic evaluation and re-anchoring guarantees for piecewise-stationary environments.
Yi-Hang Chen, Yu-Xiang Chen, Yuxuan Huang et al.· 0 citations
Large language models are increasingly deployed as interacting agents, making the persistence of their action policies across repeated interaction critical for reliable multi-agent operation. We investigate whether and how agent-generated, non-binding pre-play communication ("cheap talk") increases such persistence in...
The results show that original choices, public actions, and executed outcomes must be separated when evaluating multi-agent robustness, as corrupted communication can severely and predictably degrade mutually beneficial cooperation.
Xuan-Yi Liu, Niall J. Dalton, Hairil Amin et al.· 0 citations
A framework for mechanism design with AI agents whose alignment (preferences) and capabilities (feasible actions and information) are unknown is developed and applied to stylized examples of sandbagging and an alignment--interpretability trade-off.
Dirk Bergemann, Andrew Koh, S. Morris· 2 citations
Large language model agents deployed without a central controller are often assumed to require communication to coordinate their actions. We ask what remains possible without it: when independent instances of the same model cannot communicate, can they still reason about their counterparts well enough to exceed the sta...
Deborah Sinishaw, Qi-Le Zhu, Edwin Meriaux et al.· 2 citations
