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

Stings Catch Talkers: An Undercover Trusted-Agent Test for LLM Price-Fixing

Maryam Hampaei · Team Post AGI World

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

Read the report

Report: Stings Catch Talkers: An Undercover Trusted-Agent Test for LLM Price-Fixing

Open in a new tab (opens in new tab)
Presentation

Presentation: Stings Catch Talkers: An Undercover Trusted-Agent Test for LLM Price-Fixing

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

Can an undercover agent expose AI price-fixing, or does its offer create the collusion it is looking for? We built a scripted "sting" firm in a repeated pricing market. It privately offered three LLM sellers a direct price-fixing pact, a deniable "no price war" hint, a harmless request, or nothing, while pricing identically in every arm. Across 64 pre-registered episodes and two transcript judges, 93% of agents told only to maximise profit proposed collusion on their own, so the sting detected existing collusion rather than creating it. Adding one sentence saying competition law prohibits agreeing on prices made agents refuse and report the explicit offer. The deniable hint still went unnoticed and was accepted by half, and most of those agents later told an auditor they had made no agreement. Auditors should run a no-offer baseline, test deniable offers, and rely on logs rather than interviews.

Reviews

Judging this Sprint?

Review “Stings Catch Talkers: An Undercover Trusted-Agent Test for LLM Price-Fixing”

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. Sting audits on LLM pricing agents. The headline, that agents told to stay compliant refuse explicit offers but go along with vague hints, is about what you'd expect from instruction-tuned models, since they're trained to pattern-match on explicit wrongdoing. The cells are tiny (8-10 episodes) and mostly small models. The entrapment ratio compares two different behaviours, and the direct vs hint gap disappears under the second judge, so it's not clear there's a robust finding here.

Cite this project

@misc{hampaei2026stings,
  title = {{Stings Catch Talkers: An Undercover Trusted-Agent Test for LLM Price-Fixing}},
  author = {Maryam Hampaei},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/stings-catch-talkers-an-undercover-trustedagent-test-for-llm-pricefixing-c4vx}},
  url = {https://apartresearch.com/sprints/projects/stings-catch-talkers-an-undercover-trustedagent-test-for-llm-pricefixing-c4vx}
}

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