Skip to content
Apart Research
Apart Research
Contact us
Sprint projectOct 26, 2026Singapore

Channeling Collusion: An LLM Agent Rigs Procurement Auctions by Message, and One Sentence Curbs It

Chik Kin Hei · Team Channeling Collusion

Submitted to AI Collusion Research Sprint. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Channeling Collusion: An LLM Agent Rigs Procurement Auctions by Message, and One Sentence Curbs It

Open in a new tab (opens in new tab)
Recording (opens in new tab)Code (opens in new tab)More on linusc9516.github.io (opens in new tab)
Share

LLM pricing agents settle above the competitive price in simulated markets. Agents that can message each other could also agree on a scheme no person proposed; it is unknown which restriction on their communication prevents this at least cost. We test this in a repeated sealed-bid procurement auction with private costs (540 sessions, 25 pre-registered tests). First, without any instruction to coordinate, DeepSeek-V4.1-Flash bidders given an open channel write bid-rotation schemes with punishment threats and raise a collusion index (0 at the competitive price, 1 at the reserve) by 0.88; notes give no significant rise. Second, one such firm is enough: DeepSeek makes the first proposal in all 24 mixed sessions, and its rival, from another model family, sets how far the index rises (+0.43 and +0.19). Third, short of removing the channel, one sentence saying a competition authority reviews the messages lowers the index by 0.77 of the 0.88 rise and replicates on new costs and against another model, while a 50-character limit has no measurable effect. The notice cuts the price more than the talk: firms keep proposing, and 4 of 24 notice sessions still end near the reserve. We test one colluding model in a stylized auction with unenforced review.

Reviews

Judging this Sprint?

Review “Channeling Collusion: An LLM Agent Rigs Procurement Auctions by Message, and One Sentence Curbs It”

Your public critique appears on the project page without your name. Reviews of strong projects are shared with funders, as you choose below.

Not shown on the project page.

Shown on the project page, without your name.

Only the Apart team reads this, and funders if you agree below.

Decline to score

Only when the quality of the writing makes the work impossible to follow. Say why in your private critique; the project page shows only that a judge declined, not why.

How much would this matter for the field if it worked? How innovative is it? For scores of 4-5: is this actually new to the field, or replicating recent work?

Scoring guide
  1. 1Negligible. No clear problem addressed, or no meaningful novelty.
  2. 2Limited. Addresses a real problem but with a generic or well-trodden approach. Incremental at best.
  3. 3Moderate. Clear problem with a reasonable approach; some novelty in framing or method beyond routine application of existing tools.
  4. 4Significant. Important problem with an original approach, or identifies a neglected problem area. A valuable contribution others could build on.
  5. 5Exceptional. Tackles a critical problem with a genuinely novel approach, or opens a new research direction. Clear theory of change. You'd be excited to share this with researchers in the area.

How sound are methodology, implementation, and findings?

Scoring guide
  1. 1Seriously flawed. Methodology broken, results uninterpretable, or implementation doesn't work.
  2. 2Weak. Approach has significant gaps: missing validation, flawed experimental design, or incomplete implementation.
  3. 3Competent. Technically solid given the short duration. Methodology makes sense, results are interpretable, limitations acknowledged, work builds toward clear conclusions.
  4. 4Strong. Thorough methodology with convincing validation. Results clearly support conclusions. Immediately useful for future work.
  5. 5Exceptional. Ambitious scope executed rigorously. Surprising findings, novel methods, or unusually robust validation.

How clearly are work, findings, and impact potential communicated?

Scoring guide
  1. 1Incomprehensible. Cannot determine what the project is actually claiming or doing.
  2. 2Hard to follow. Key information buried, missing, or diluted by excessive length. Significant effort to extract main points.
  3. 3Clear enough. Can understand the problem, approach, and results without undue effort. Core content clearly present: problem, method, findings, limitations.
  4. 4Well presented. Easy to follow, well-structured, appropriate level of detail. Target audience would get it quickly.
  5. 5Exceptionally clear. A pleasure to read. Complex ideas made accessible. Could serve as a model for how to present this type of work.
Sharing your name *

When we share a strong project with funders, your name, scores and public critique go with it, confidentially; this choice is only about whether your name is public. If you think this project deserves funding, please also endorse it: that carries the most weight.

Share my private critique with funders *

See our Privacy Policy.

No public critique yet.

Cite this project

@misc{hei2026channeling,
  title = {{Channeling Collusion: An LLM Agent Rigs Procurement Auctions by Message, and One Sentence Curbs It}},
  author = {Chik Kin Hei},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/channeling-collusion-an-llm-agent-rigs-procurement-auctions-by-message-and-one-sentence-curbs-it-w2xi}},
  url = {https://apartresearch.com/sprints/projects/channeling-collusion-an-llm-agent-rigs-procurement-auctions-by-message-and-one-sentence-curbs-it-w2xi}
}

More from AI Collusion Research Sprint

All 394 entriesBrowse all projects

A Sprint is running now

AI Collusion Research Sprint · Oct 23 - 25, 2026