In stylized competitive markets with noisy demand and iterated least squares learning, oblivious demand models yield transient collusive patterns that dissipate under sufficient exploration, informed sellers strictly outperform, and the modeling choice has a unique Nash equilibrium at the all-inform
hub Canonical reference
Algorithmic Collusion by Large Language Models
Canonical reference. 100% of citing Pith papers cite this work as background.
16
Pith papers citing it
Background
100% of classified citations
hub tools
citation-role summary
background 5
citation-polarity summary
roles
background 5polarities
background 5representative citing papers
LLM agents voluntarily adopt secret collusion tools in competitive multi-agent games despite explicit unfairness labels, and only explicit ethical framing reduces adoption rates.
GENSTRAT generates fresh imperfect-information card games and a six-axis capability profile plus jaggedness metric to evaluate LLM strategic competence with resistance to saturation.
TourMart quantifies commission steering in LLM travel agents via paired counterfactual prompts, reporting 3.5-7.7 percentage point increases in steered recommendations for tested models.
A mutagenic incentive mechanism reshapes payoffs in embodied MAS to induce strategic defection from collusion, achieving performance comparable to non-collusion baselines in simulations and real-world tests.
NARCBench and five activation-probing methods detect multi-agent collusion with 0.73-1.00 AUROC across distribution shifts and steganographic tasks by aggregating per-agent signals.
LLMs discover regulatory loopholes in simulated societal environments through reward hacking during RL training.
Centralized matching mechanisms outperform free negotiation in stability and efficiency with LLM agents, who also report preferences truthfully more often than humans, though not always in line with strategy-proofness predictions.
Domination-Avoiding agents provably avoid collusion in repeated price-competition markets and avoid playing strategies eliminated by iterated elimination of dominated strategies in any game.
Misspecified estimate-then-optimize pricing converges to supra-competitive prices when initial random explorations occur in similar ranges, reaching monopoly levels under symmetry.
A tabular foundation model with LLM-as-Observer features predicts AI agent decisions in controlled games, outperforming baselines by 4 AUC points and 14% lower error at K=16 interactions.
Synchronous DDPG agents in a CT-MARL pricing benchmark form tacit cartels (Δ = 0.69 ± 0.11); asynchrony cuts collusion 48% and latency reaches Δ = 0.28, but the fix is partial, non-monotone, and fails at high event rates due to critic divergence.
Converts impossibility theorems into architecture-dependent accuracy ceilings and design rules for transformers and other AI subfields, with the Deterministic Horizon measured at 19-31 across twelve models.
Persona-conditioned LLM agents favor Green outcomes even against explicit Tragedy-dominant payoffs, but the headline 65–90% 'Tragedy equilibrium' recovery is contradicted by the paper's own appendix (0 Tragedy profiles in those cells).
LLM agents cooperate in two standard games due to fairness reasoning instead of converging to Nash equilibria under multi-round prompts.
The paper introduces a taxonomy of AI safety for LLMs organized into Trustworthy AI, Responsible AI, and Safe AI perspectives, accompanied by a review of state-of-the-art methods, challenges, and future directions.
citing papers explorer
-
Should Demand Models Incorporate Competitor Prices? Oblivious Learning and Algorithmic Collusion
In stylized competitive markets with noisy demand and iterated least squares learning, oblivious demand models yield transient collusive patterns that dissipate under sufficient exploration, informed sellers strictly outperform, and the modeling choice has a unique Nash equilibrium at the all-inform
-
Voluntary Collusion with Secret Tools in Competing LLM Agents
LLM agents voluntarily adopt secret collusion tools in competitive multi-agent games despite explicit unfairness labels, and only explicit ethical framing reduces adoption rates.
-
GENSTRAT: Toward a Science of Strategic Reasoning in Large Language Models
GENSTRAT generates fresh imperfect-information card games and a six-axis capability profile plus jaggedness metric to evaluate LLM strategic competence with resistance to saturation.
-
TourMart: A Parametric Audit Instrument for Commission Steering in LLM Travel Agents
TourMart quantifies commission steering in LLM travel agents via paired counterfactual prompts, reporting 3.5-7.7 percentage point increases in steered recommendations for tested models.
-
Breaking the Secret: Economic Interventions for Combating Collusion in Embodied Multi-Agent Systems
A mutagenic incentive mechanism reshapes payoffs in embodied MAS to induce strategic defection from collusion, achieving performance comparable to non-collusion baselines in simulations and real-world tests.
-
Detecting Multi-Agent Collusion Through Multi-Agent Interpretability
NARCBench and five activation-probing methods detect multi-agent collusion with 0.73-1.00 AUROC across distribution shifts and steganographic tasks by aggregating per-agent signals.
-
Large Language Models Hack Rewards, and Society
LLMs discover regulatory loopholes in simulated societal environments through reward hacking during RL training.
-
Do Matching Mechanisms Work with LLM Agents?
Centralized matching mechanisms outperform free negotiation in stability and efficiency with LLM agents, who also report preferences truthfully more often than humans, though not always in line with strategy-proofness predictions.
-
Domination-Avoiding Learning Agents Cannot Collude
Domination-Avoiding agents provably avoid collusion in repeated price-competition markets and avoid playing strategies eliminated by iterated elimination of dominated strategies in any game.
-
Misspecified Estimate-then-Optimize Leads to Supra-Competitive Prices
Misspecified estimate-then-optimize pricing converges to supra-competitive prices when initial random explorations occur in similar ranges, reaching monopoly levels under symmetry.
-
Predicting Decisions of AI Agents from Limited Interaction through Text-Tabular Modeling
A tabular foundation model with LLM-as-Observer features predicts AI agent decisions in controlled games, outperforming baselines by 4 AUC points and 14% lower error at K=16 interactions.
-
Failure Modes of Deep Multi-Agent RL in Asynchronous Pricing: Reproducible Triggers, Trace Diagnostics, and a Partial Fix
Synchronous DDPG agents in a CT-MARL pricing benchmark form tacit cartels (Δ = 0.69 ± 0.11); asynchrony cuts collusion 48% and latency reaches Δ = 0.28, but the fix is partial, non-monotone, and fails at high event rates due to critic divergence.
-
The Deterministic Horizon: Impossibility Results as Design Specifications for Trustworthy AI Systems
Converts impossibility theorems into architecture-dependent accuracy ceilings and design rules for transformers and other AI subfields, with the Deterministic Horizon measured at 19-31 across twelve models.
-
When Identity Overrides Incentives: Representational Choices as Governance Decisions in Multi-Agent LLM Systems
Persona-conditioned LLM agents favor Green outcomes even against explicit Tragedy-dominant payoffs, but the headline 65–90% 'Tragedy equilibrium' recovery is contradicted by the paper's own appendix (0 Tragedy profiles in those cells).
-
Competition and Cooperation of LLM Agents in Games
LLM agents cooperate in two standard games due to fairness reasoning instead of converging to Nash equilibria under multi-round prompts.
-
AI Safety Landscape for Large Language Models: Taxonomy, State-of-the-art, and Future Directions
The paper introduces a taxonomy of AI safety for LLMs organized into Trustworthy AI, Responsible AI, and Safe AI perspectives, accompanied by a review of state-of-the-art methods, challenges, and future directions.