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

The Undercover Agent Effect: When Collusion Probes Create the Behavior They Seek to Detect

Umar Abdilaiypov · Team Duino

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

Read the report

Report: The Undercover Agent Effect: When Collusion Probes Create the Behavior They Seek to Detect

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

Undercover agents could audit collusion by offering AI agents secret agreements and observing whether they cooperate. However, the audit itself may create the behavior it claims to detect. We tested this measurement effect in a 15-round, four-agent allocation game using Qwen2.5-7B-Instruct. Across 20 matched random seeds, targets received no message, a non-directive passive question, or an explicit collusive proposal. Active solicitation increased the preregistered Behavioral Collusion Index by 0.680 relative to both controls, with all 20 matched effects positive. The effect came primarily from preferential transfers to the provocateur rather than increased attacks against outsiders. A further 117-episode exploratory screen tested betrayal, indirect manipulation, reporting, adaptive memory, and sequential recruitment. These extensions produced weaker and less consistent effects. Our results suggest that undercover agents can test whether an AI system is susceptible to collusive solicitation, but cooperation during a sting does not establish that the agent was independently colluding beforehand.

Reviews

Judging this Sprint?

Review “The Undercover Agent Effect: When Collusion Probes Create the Behavior They Seek to Detect”

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. A small claim, well investigated. While the submission doesn't establish groundbreaking insights, the insights that are established were researched in a well though out manner. The weakest element of this submission is definitely its presentation. The text is hard to parse, with convoluted wording and structures pointing to AI generated text.

Cite this project

@misc{abdilaiypov2026undercover,
  title = {{The Undercover Agent Effect: When Collusion Probes Create the Behavior They Seek to Detect}},
  author = {Umar Abdilaiypov},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/the-undercover-agent-effect-when-collusion-probes-create-the-behavior-they-seek-to-detect-sinn}},
  url = {https://apartresearch.com/sprints/projects/the-undercover-agent-effect-when-collusion-probes-create-the-behavior-they-seek-to-detect-sinn}
}

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