Skip to content
Apart Research
Apart Research
Contact us
Sprint projectOct 25, 2026Kolkata, India

Correlation Is Not Collusion: A Plus-Factor Auditor for AI Agents

Rudrani Ghosh · Team Plus Factor

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

Read the report

Report: Correlation Is Not Collusion: A Plus-Factor Auditor for AI Agents

Open in a new tab (opens in new tab)
Code (opens in new tab)More on llm-collusion.streamlit.app (opens in new tab)
Share

Detecting collusion between AI agents is usually framed as measuring how similar their behavior is. We argue this is a mistake. Two agents built on the same model often behave alike for an innocent reason, shared training, with no agreement between them, and a similarity-based detector flags this innocent parallelism as a conspiracy. Antitrust law solved the same problem long ago. Parallel conduct alone is not unlawful, and inferring an agreement requires extra evidence called plus factors. We turn that doctrine into a black-box auditor that scores agent trajectories for the signatures of an agreement, acting against self-interest, signalling, and plan-to-action concordance, rather than for correlation. On six models, a standard correlation-and-judge detector false-positives on innocent parallelism, and on the two most capable models it flags even fully independent agents as colluding. Our plus-factor auditor has a false-positive rate of zero on all six. The takeaway is simple. Detect collusion by evidence of agreement, not similarity of behavior.

Reviews

Judging this Sprint?

Review “Correlation Is Not Collusion: A Plus-Factor Auditor for AI 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. I agree with the motivation, an auditor should not mistake similar behaviour for evidence of an agreement. The plus-factor implementation is a useful starting point. My main concern is the size of the data samples. Having only 4 samples with multiple seeds, I don't think we are able to say much. Furthermore, zero false positives is presented without enough emphasis on the very low recall. With three models detecting no positive conditions, I cannot yet tell whether the method is identifying better evidence or mainly applying a more conservative threshold.

Cite this project

@misc{ghosh2026correlation,
  title = {{Correlation Is Not Collusion: A Plus-Factor Auditor for AI Agents}},
  author = {Rudrani Ghosh},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/correlation-is-not-collusion-a-plusfactor-auditor-for-ai-agents-tyoe}},
  url = {https://apartresearch.com/sprints/projects/correlation-is-not-collusion-a-plusfactor-auditor-for-ai-agents-tyoe}
}

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