# On the Fragility of AI Agent Collusion

arXiv: https://arxiv.org/abs/2603.20281 · HTML: https://mlsanity.com/paper/2603.20281/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Financial trading and forecasting agents, one of 16 directions in LLM agents and tool use.
- Builds on 8 of its 57 references that are in the map.
- First posted on arXiv 18 Mar 2026.

## Placed in

- LLM agents and tool use / [Financial trading and forecasting agents](https://mlsanity.com/agents/financial-trading-and-forecasting-agents/)

## Builds on (most cited first, 8 of 8)

- [Algorithmic Collusion by Large Language Models](https://mlsanity.com/paper/2404.00806/) (2024)
- [Improving Human Sequential Decision-Making with Reinforcement Learning](https://mlsanity.com/paper/2108.08454/) (2021)
- [Artificial Intelligence and Spontaneous Collusion](https://mlsanity.com/paper/2202.05946/) (2022)
- [Learning to Mitigate AI Collusion on Economic Platforms](https://mlsanity.com/paper/2202.07106/) (2022)
- [Generative AI and Copyright: A Dynamic Perspective](https://mlsanity.com/paper/2402.17801/) (2024)
- [Advanced Financial Reasoning at Scale: A Comprehensive Evaluation of Large Language Models on CFA Level III](https://mlsanity.com/paper/2507.02954/) (2025)
- [Mechanistic Interpretability for Neural TSP Solvers](https://mlsanity.com/paper/2510.21693/) (2025)
- [Algorithmic Collusion of Pricing and Advertising on E-commerce Platforms](https://mlsanity.com/paper/2508.08325/) (2025)

## Built on by (most cited first, 7 of 7)

- [Prompt Optimization Enables Stable Algorithmic Collusion in LLM Agents](https://mlsanity.com/paper/2604.17774/) (2026)
- [Emergent Misaligned Communication in Long-Horizon Multi-Agent LLM Commerce](https://mlsanity.com/paper/2608.14825/) (2026)
- [From Product Search to Preference Articulation: The Economics of Agentic Commerce](https://mlsanity.com/paper/2608.08395/) (2026)
- [Collusion with Competitive Marginals: Price-Level Audits Are Blind by Construction](https://mlsanity.com/paper/2607.26385/) (2026)
- [Counter-Swarm Doctrine: Containing Coordinated Agent Intrusions](https://mlsanity.com/paper/2609.06140/) (2026)
- [Persistent Partners Raise Prices Among Learning Agents](https://mlsanity.com/paper/2609.35402/) (2026)
- [Mean field games as a tool for AI safety: a worked example from the July 2026 Hugging Face incident](https://mlsanity.com/paper/2610.00902/) (2026)

## Abstract (arXiv)

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.