Position: Collusion Risks Among AI Reasoning Agents Justify Certification Requirements for Making Market Decisions

Matthew Riemer, Tommaso Tosato, Amin Memarian, Maximilian Puelma Touzel, Glen Berseth, Irina Rish, Guillaume Dumas
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:171880-171907, 2026.

Abstract

This position paper argues that AI agents with chain-of-thought reasoning capabilities are predisposed to exhibit collusive behavior and should be required to obtain behavioral certification before making decisions that affect economic markets. This is because integrating these agents into society could collapse the legal evidentiary distinction between competition and collusion among independent firms without eroding the economic harm distinction. Experiments with DeepSeek-R1 agents in the Bertrand oligopoly pricing domain reveal a tendency towards tacit collusion that persists even when humans prompt the agents not to collude. We further show that the chain-of-thought of these agents can be steered toward either extremely collusive or highly competitive behavior in a way that is not semantically detectable by another LLM analyzing the reasoning traces. As a result, deploying reasoning agents for market decisions leads to collusive economic outcomes without any evidence of conspiracy or intent. Thus, certification based on observed behavior in representative situations is necessary to prevent collusion. We provide preliminary evidence that such agents can be steered in a generalizable way toward efficient competitive equilibria. However, developing a comprehensive behavioral certification will be required before these models can be deployed in real-world markets while ensuring their stability and efficiency.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-riemer26a, title = {Position: Collusion Risks Among {AI} Reasoning Agents Justify Certification Requirements for Making Market Decisions}, author = {Riemer, Matthew and Tosato, Tommaso and Memarian, Amin and Puelma Touzel, Maximilian and Berseth, Glen and Rish, Irina and Dumas, Guillaume}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {171880--171907}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/riemer26a/riemer26a.pdf}, url = {https://proceedings.mlr.press/v306/riemer26a.html}, abstract = {This position paper argues that AI agents with chain-of-thought reasoning capabilities are predisposed to exhibit collusive behavior and should be required to obtain behavioral certification before making decisions that affect economic markets. This is because integrating these agents into society could collapse the legal evidentiary distinction between competition and collusion among independent firms without eroding the economic harm distinction. Experiments with DeepSeek-R1 agents in the Bertrand oligopoly pricing domain reveal a tendency towards tacit collusion that persists even when humans prompt the agents not to collude. We further show that the chain-of-thought of these agents can be steered toward either extremely collusive or highly competitive behavior in a way that is not semantically detectable by another LLM analyzing the reasoning traces. As a result, deploying reasoning agents for market decisions leads to collusive economic outcomes without any evidence of conspiracy or intent. Thus, certification based on observed behavior in representative situations is necessary to prevent collusion. We provide preliminary evidence that such agents can be steered in a generalizable way toward efficient competitive equilibria. However, developing a comprehensive behavioral certification will be required before these models can be deployed in real-world markets while ensuring their stability and efficiency.} }
Endnote
%0 Conference Paper %T Position: Collusion Risks Among AI Reasoning Agents Justify Certification Requirements for Making Market Decisions %A Matthew Riemer %A Tommaso Tosato %A Amin Memarian %A Maximilian Puelma Touzel %A Glen Berseth %A Irina Rish %A Guillaume Dumas %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-riemer26a %I PMLR %P 171880--171907 %U https://proceedings.mlr.press/v306/riemer26a.html %V 306 %X This position paper argues that AI agents with chain-of-thought reasoning capabilities are predisposed to exhibit collusive behavior and should be required to obtain behavioral certification before making decisions that affect economic markets. This is because integrating these agents into society could collapse the legal evidentiary distinction between competition and collusion among independent firms without eroding the economic harm distinction. Experiments with DeepSeek-R1 agents in the Bertrand oligopoly pricing domain reveal a tendency towards tacit collusion that persists even when humans prompt the agents not to collude. We further show that the chain-of-thought of these agents can be steered toward either extremely collusive or highly competitive behavior in a way that is not semantically detectable by another LLM analyzing the reasoning traces. As a result, deploying reasoning agents for market decisions leads to collusive economic outcomes without any evidence of conspiracy or intent. Thus, certification based on observed behavior in representative situations is necessary to prevent collusion. We provide preliminary evidence that such agents can be steered in a generalizable way toward efficient competitive equilibria. However, developing a comprehensive behavioral certification will be required before these models can be deployed in real-world markets while ensuring their stability and efficiency.
APA
Riemer, M., Tosato, T., Memarian, A., Puelma Touzel, M., Berseth, G., Rish, I. & Dumas, G.. (2026). Position: Collusion Risks Among AI Reasoning Agents Justify Certification Requirements for Making Market Decisions. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:171880-171907 Available from https://proceedings.mlr.press/v306/riemer26a.html.

Related Material