We revisit this scenario as an experimental case study in Section 6.

Recent advancements in generative AI suggest the potential for large-scale interaction between autonomous agents and humans across platforms such as the internet. While such interactions could foster productive cooperation, the ability of AI agents to circumvent security oversight raises critical multi-agent security problems, particularly in the form of unintended information sharing or undesirable coordination. In our work, we establish the subfield of secret collusion, a form of multi-agent deception, in which two or more agents employ steganographic methods to conceal the true nature of their interactions, be it communicative or otherwise, from oversight. We propose a formal threat model for AI agents communicating steganographically and derive rigorous theoretical insights about the capacity and incentives of large language models (LLMs) to perform secret collusion, in addition to the limitations of threat mitigation measures. We complement our findings with empirical evaluations demonstrating rising steganographic capabilities in frontier single and multi-agent LLM setups and examining potential scenarios where collusion may emerge, revealing limitations in countermeasures such as monitoring, paraphrasing, and parameter optimization. Our work is the first to formalize and investigate secret collusion among frontier foundation models, identifying it as a critical area in AI Safety and outlining a comprehensive research agenda to mitigate future risks of collusion between generative AI systems.

An oligopoly is a market in which the price of goods is controlled by a few firms. Cournot introduced the simplest game-theoretic model of oligopoly, where profit-maximizing behavior of each firm results in market failure. Furthermore, when the Cournot oligopoly game is infinitely repeated, firms can tacitly collude to monopolize the market. Such tacit collusion is realized by the same mechanism as direct reciprocity in the repeated prisoner's dilemma game, where mutual cooperation can be realized whereas defection is favorable for both prisoners in a one-shot game. Recently, in the repeated prisoner's dilemma game, a class of strategies called zero-determinant strategies attracts much attention in the context of direct reciprocity. Zero-determinant strategies are autocratic strategies which unilaterally control payoffs of players by enforcing linear relationships between payoffs. There were many attempts to find zero-determinant strategies in other games and to extend them so as to apply them to broader situations. In this paper, first, we show that zero-determinant strategies exist even in the repeated Cournot oligopoly game, and that they are quite different from those in the repeated prisoner's dilemma game. Especially, we prove that a fair zero-determinant strategy exists, which is guaranteed to obtain the average payoff of the opponents. Second, we numerically show that the fair zero-determinant strategy can be used to promote collusion when it is used against an adaptively learning player, whereas it cannot promote collusion when it is used against two adaptively learning players. Our findings elucidate some negative impact of zero-determinant strategies in the oligopoly market.

Abstract--The integration of large language models (LLMs) into healthcare IoT systems promises faster decisions and improved medical support. LLMs are also deployed as multi-agent teams to assist AI doctors by debating, voting, or advising on decisions. However, when multiple assistant agents interact, coordinated adversaries can collude to create false consensus, pushing an AI doctor toward harmful prescriptions. We develop an experimental framework with scripted and unscripted doctor agents, adversarial assistants, and a verifier agent that checks decisions against clinical guidelines. Using 50 representative clinical questions, we find that collusion drives the Attack Success Rate (ASR) and Harmful Recommendation Rates (HRR) up to 100% in unprotected systems. This work provides the first systematic evidence of collusion risk in AI healthcare and demonstrates a practical, lightweight defence that ensures guideline fidelity. Artificial intelligence (AI) is increasingly integrated into healthcare IoT systems, supporting tasks such as remote patient monitoring, diagnosis, and treatment recommendations. In this setting, ensuring the security and trustworthiness of AI decisions is critical, since medical errors caused by unsafe recommendations can severely harm patients [1]. However, AI doctors and LLM-based clinical decision agents face multiple vulnerabilities.

AI agents are increasingly deployed in ecosystems where they repeatedly interact not only with each other but also with humans. In this work, we study these human-AI ecosystems from a theoretical perspective, focusing on the classical framework of repeated pricing games. In our stylized model, the AI agents play equilibrium strategies, and one or more humans manually perform the pricing task instead of adopting an AI agent, thereby defecting to a no-regret strategy. Motivated by how populations of AI agents can sustain supracompetitive prices, we investigate whether high prices persist under such defections. Our main finding is that even a single human defection can destabilize collusion and drive down prices, and multiple defections push prices even closer to competitive levels. We further show how the nature of collusion changes under defection-aware AI agents. Taken together, our results characterize when algorithmic collusion is fragile--and when it persists--in mixed ecosystems of AI agents and humans.

We study whether large language models acting as autonomous bidders can tacitly collude by coordinating when to accept platform posted payouts in repeated Dutch auctions, without any communication. We present a minimal repeated auction model that yields a simple incentive compatibility condition and a closed form threshold for sustainable collusion for subgame-perfect Nash equilibria. In controlled simulations with multiple language models, we observe systematic supra-competitive prices in small auction settings and a return to competitive behavior as the number of bidders in the market increases, consistent with the theoretical model. We also find LLMs use various mechanisms to facilitate tacit coordination, such as focal point acceptance timing versus patient strategies that track the theoretical incentives. The results provide, to our knowledge, the first evidence of bidder side tacit collusion by LLMs and show that market structure levers can be more effective than capability limits for mitigation.

Recent findings reveal that even simple pricing algorithms can make things more expensive. Imagine a town with two widget merchants. Customers prefer cheaper widgets, so the merchants must compete to set the lowest price. Unhappy with their meager profits, they meet one night in a smoke-filled tavern to discuss a secret plan: If they raise prices together instead of competing, they can both make more money. But that kind of intentional price-fixing, called collusion, has long been illegal.

As algorithms increasingly mediate competitive decision-making, their influence extends beyond individual outcomes to shaping strategic market dynamics. In two preregistered experiments, we examined how algorithmic advice affects human behavior in classic economic games with unique, non-collusive, and analytically traceable equilibria. In Experiment 1 (N = 107), participants played a Bertrand price competition with individualized or collective algorithmic recommendations. Initially, collusively upward-biased advice increased prices, particularly when individualized, but prices gradually converged toward equilibrium over the course of the experiment. However, participants avoided setting prices above the algorithm's recommendation throughout the experiment, suggesting that advice served as a soft upper bound for acceptable prices. In Experiment 2 (N = 129), participants played a Cournot quantity competition with equilibrium-aligned or strategically biased algorithmic recommendations. Here, individualized equilibrium advice supported stable convergence, whereas collusively downward-biased advice led to sustained underproduction and supracompetitive profits - hallmarks of tacit collusion. In both experiments, participants responded more strongly and consistently to individualized advice than collective advice, potentially due to greater perceived ownership of the former. These findings demonstrate that algorithmic advice can function as a strategic signal, shaping coordination even without explicit communication. The results echo real-world concerns about algorithmic collusion and underscore the need for careful design and oversight of algorithmic decision-support systems in competitive environments.

When online sellers use AI learning algorithms to automatically compete on e-commerce platforms, there is concern that they will learn to coordinate on higher than competitive prices. However, this concern was primarily raised in single-dimension price competition. We investigate whether this prediction holds when sellers make pricing and advertising decisions together, i.e., two-dimensional decisions. We analyze competition in multi-agent reinforcement learning, and use a large-scale dataset from Amazon.com to provide empirical evidence. We show that when consumers have high search costs, learning algorithms can coordinate on prices lower than competitive prices, facilitating a win-win-win for consumers, sellers, and platforms. This occurs because algorithms learn to coordinate on lower advertising bids, which lower advertising costs, leading to lower prices and enlarging demand on the platform. We also show that our results generalize to any learning algorithm that uses exploration of price and advertising bids. Consistent with our predictions, an empirical analysis shows that price levels exhibit a negative interaction between estimated consumer search costs and algorithm usage index. We analyze the platform's strategic response and find that reserve price adjustments will not increase platform profits, but commission adjustments will, while maintaining the beneficial outcomes for both sellers and consumers.