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

Making Multi-Agent Collusion Auditable under ISO/IEC 42001

Sourav Mondal, Suchisubhra Das

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

Read the report

Report: Making Multi-Agent Collusion Auditable under ISO/IEC 42001

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

Presentation: Making Multi-Agent Collusion Auditable under ISO/IEC 42001

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

Multi-agent AI deployments increasingly use an evaluator agent to check a task agent's outputs. When the two share a reward, model lineage, data store, or communication channel, the evaluator can inflate grades — a collusion failure that no AI governance standard currently tells organisations how to evidence against. We present SentinelAIMS, an ISO/IEC 42001 management-system tool extended with a multi-agent collusion overlay. Applying segregation-of-duties practice from privileged access management, we map three collusion risks — evaluator–evaluatee coordination, shared-context contamination, and covert signalling — to specific Annex A controls (A.3.2, A.4.3–A.4.5, A.6.2.4, A.6.2.6–A.6.2.8) and ISO/IEC 27001 companion controls, and define the evidence artifact an auditor would require for each. A built-in scenario (JudgeNet grading TriageFlow) illustrates the workflow with hand-seeded data: without controls, +21.5 average grade inflation; with distinct identities and separate infrastructure, deviations within ±3. These figures are illustrative, not experimental, and the mapping has not yet been reviewed by a certified auditor. Our takeaway: ISO 42001 cannot detect collusion, but it can make collusion defences auditable today — a practical governance floor on which detection research can build.

Reviews

Judging this Sprint?

Review “Making Multi-Agent Collusion Auditable under ISO/IEC 42001”

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. Innovative and pragmatic draft for a policy to govern multi-agent collusion. Given more time and feedback from experts this work could "get there" and help us reduce related risks to collusion.

Cite this project

@misc{mondal2026making,
  title = {{Making Multi-Agent Collusion Auditable under ISO/IEC 42001}},
  author = {Sourav Mondal and Suchisubhra Das},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/making-multiagent-collusion-auditable-under-isoiec-42001-k5kt}},
  url = {https://apartresearch.com/sprints/projects/making-multiagent-collusion-auditable-under-isoiec-42001-k5kt}
}

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