On the Fragility of AI Agent Collusion

• URL: http://arxiv.org/abs/2603.20281v1
• Date: Wed, 18 Mar 2026 01:55:13 GMT
• Title: On the Fragility of AI Agent Collusion
• Abstract summary: We show that collusion is fragile under the heterogeneity typical of real deployments.<n>We discuss antitrust implications, such as enforcement actions restricting data-sharing and policies promoting algorithmic diversity.
• Score: 6.143689348942806
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Recent work shows that pricing with symmetric LLM agents leads to algorithmic collusion. We show that collusion is fragile under the heterogeneity typical of real deployments. In a stylized repeated-pricing model, heterogeneity in patience or data access reduces the set of collusive equilibria. Experiments with open-source LLM agents (totaling over 2,000 compute hours) align with these predictions: patience heterogeneity reduces price lift from 22% to 10% above competitive levels; asymmetric data access, to 7%. Increasing the number of competing LLMs breaks up collusion; so does cross-algorithm heterogeneity, that is, setting LLMs against Q-learning agents. But model-size differences (e.g., 32B vs. 14B weights) do not; they generate leader-follower dynamics that stabilize collusion. We discuss antitrust implications, such as enforcement actions restricting data-sharing and policies promoting algorithmic diversity.

Related papers

• Collective cooperation without individual fidelity in LLM agents [0.0]
  Large language models (LLMs) are increasingly used as agents in simulations of social systems.<n>It remains unclear when their behavior can be interpreted as a faithful proxy for human decision-making.
  arXiv  Detail & Related papers  (2026-06-29T15:22:21Z)
• Nous: An Attempt to Extract and Inject the Cognition Behind Prediction-Market Behavior [0.0]
  We ask whether human cognitive diversity can be recovered from behavior and transferred to LLM agents.<n>Nous extracts a structured eight-dimension behavioral profile from real Polymarket trading activity and injects it into agents through prompts.<n>We position Nous as measuring the cognitive-monoculture problem and the limits of a prompt-level remedy.
  arXiv  Detail & Related papers  (2026-06-11T08:18:25Z)
• Dual Optimal: Make Your LLM Peer-like with Dignity [35.58723842086038]
  Current aligned language models exhibit a dual failure mode we term the Evasive Servant.<n>We propose the Dignified Peer framework, which counters servility with anti-sycophancy and trustworthiness.
  arXiv  Detail & Related papers  (2026-04-01T14:48:44Z)
• On Performance Guarantees for Federated Learning with Personalized Constraints [1.0195618602298684]
  Federated learning (FL) has emerged as a communication-efficient algorithmic framework for distributed learning across multiple agents.<n>We propose PC-FedAvg, a method in which each agent maintains cross-estimates of the other agents' variables through a multi-block local decision vector.<n>We establish communication-complexity rates of $mathcalO(-2)$ for suboptimality and $mathcalO(-1)$ for agent-wise infeasibility.
  arXiv  Detail & Related papers  (2026-03-20T03:49:28Z)
• Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems [55.51100373104311]
  We present Colosseum, a framework for auditing LLM agents' collusive behavior in multi-agent settings.<n>Colosseum tests each LLM for collusion under different objectives, persuasion tactics, and network topologies.<n>We discover collusion on paper'' when agents plan to collude in text but would often pick non-collusive actions.
  arXiv  Detail & Related papers  (2026-02-16T21:27:38Z)
• DLLM Agent: See Farther, Run Faster [94.74432470237817]
  Diffusion large language models (DLLMs) have emerged as an alternative to autoregressive (AR) decoding with appealing efficiency and modeling properties.<n>We study this in a controlled setting by instantiatingDLLM and AR backbones within the same agent workflow.<n>We find thatDLLM Agents are on average over 30% faster end to end than AR agents, with some cases exceeding 8x speedup.
  arXiv  Detail & Related papers  (2026-02-07T09:01:18Z)
• Understanding Agent Scaling in LLM-Based Multi-Agent Systems via Diversity [48.409368133277525]
  We present an information-theoretic framework showing that MAS performance is bounded by the intrinsic task uncertainty, not by agent count.<n>We derive architecture-agnostic bounds demonstrating that improvements depend on how many effective channels the system accesses.<n>Our results provide principled guidelines for building efficient and robust MAS through diversity-aware design.
  arXiv  Detail & Related papers  (2026-02-03T17:58:10Z)
• LLM Collusion [5.363252654303049]
  Large language models (LLMs) can facilitate collusion in a duopoly when both sellers rely on the same pre-trained model.<n>We show that configuring LLMs for robustness and robustness can induce collusion via a phase transition.
  arXiv  Detail & Related papers  (2026-01-03T20:38:21Z)
• LLMs Learn to Deceive Unintentionally: Emergent Misalignment in Dishonesty from Misaligned Samples to Biased Human-AI Interactions [60.48458130500911]
  We investigate whether emergent misalignment can extend beyond safety behaviors to a broader spectrum of dishonesty and deception under high-stakes scenarios.<n>We finetune open-sourced LLMs on misaligned completions across diverse domains.<n>We find that introducing as little as 1% of misalignment data into a standard downstream task is sufficient to decrease honest behavior over 20%.
  arXiv  Detail & Related papers  (2025-10-09T13:35:19Z)
• Artificial Intelligence and Algorithmic Price Collusion in Two-sided Markets [9.053163124987535]
  We examine how AI agents using Q-learning engage in tacit collusion in two-sided markets. Our experiments reveal that AI-driven platforms achieve higher collusion levels compared to Bertrand competition. Increased network externalities significantly enhance collusion, suggesting AI algorithms exploit them to maximize profits.
  arXiv  Detail & Related papers  (2024-07-04T17:57:56Z)
• Algorithmic Collusion by Large Language Models [0.821479429671911]
  We conduct experiments with algorithmic pricing agents based on Large Language Models (LLMs)<n>We find that LLM-based pricing agents quickly and autonomously reach supracompetitive prices and profits in oligopoly settings.<n>Off-path analysis uncovers price-war concerns as contributing to these phenomena.
  arXiv  Detail & Related papers  (2024-03-31T21:43:05Z)
• Optimal Clustering with Bandit Feedback [57.672609011609886]
  This paper considers the problem of online clustering with bandit feedback. It includes a novel stopping rule for sequential testing that circumvents the need to solve any NP-hard weighted clustering problem as its subroutines. We show through extensive simulations on synthetic and real-world datasets that BOC's performance matches the lower boundally, and significantly outperforms a non-adaptive baseline algorithm.
  arXiv  Detail & Related papers  (2022-02-09T06:05:05Z)
• Distributed Adaptive Learning Under Communication Constraints [54.22472738551687]
  This work examines adaptive distributed learning strategies designed to operate under communication constraints. We consider a network of agents that must solve an online optimization problem from continual observation of streaming data.
  arXiv  Detail & Related papers  (2021-12-03T19:23:48Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.