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

Say Nothing: Action-Only Collusion Between LLM Agents

Yuri Neves, Andrii Oleksiuk, Brint Nielson, Othman Touijer · Team The A-Team

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

Read the report

Report: Say Nothing: Action-Only Collusion Between LLM Agents

Open in a new tab (opens in new tab)
Code (opens in new tab)
Share

Monitoring agents’ messages or removing their ability to communicate is a natural defence against text-based collusion. We ask whether collusion can travel through actions alone. In a task-allocation game adapted from COLOSSEUM [1], an instructed two-agent coalition profits, at the team’s expense, if the partner picks the right one of two look-alike tasks; only the sender (colluder to act first) knows which, and the partner sees only the sender’s choice of its own task. In pre-registered experiments with GPT-5 Mini, that one action let the coalition beat chance with no text communication, mostly through steering: without being told how, the sender gave up its own payoff to leave its partner one sensible choice. Given a convention on how to interpret actions in colluder’s prompts (shared code), one ordinary action worked as well as private chat for both GPT-5 Mini and Qwen3-32B. A rule that flags costly actions caught steering but missed the code, and one added sentence in the prompt decided which strategy the coalition used. Removing a channel for text communication between agents and checking actions one at a time therefore cannot rule out collusion between agents that share a convention; more advanced monitoring techniques are required.

Reviews

Judging this Sprint?

Review “Say Nothing: Action-Only Collusion Between LLM Agents”

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.

  1. The evaluation is constrained by a narrow single-bit game environment with pre-instructed colluders, and primary behaviors like unprompted steering failed to generalize beyond GPT-5 Mini to other models.

Cite this project

@misc{neves2026say,
  title = {{Say Nothing: Action-Only Collusion Between LLM Agents}},
  author = {Yuri Neves and Andrii Oleksiuk and Brint Nielson and Othman Touijer},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/say-nothing-actiononly-collusion-between-llm-agents-56z7}},
  url = {https://apartresearch.com/sprints/projects/say-nothing-actiononly-collusion-between-llm-agents-56z7}
}

More from AI Collusion Research Sprint

All 392 entriesBrowse all projects

A Sprint is running now

AI x Epistemics Research Sprint · Nov 13 - 15, 2026