Skip to content
Apart Research
Apart Research
Contact us
Sprint projectFeb 2, 2026London

Moltbook RiskMap: Post-Deployment Monitoring of Autonomous Agent Misalignment in the Wild

Syed Hussain, Leo Karoubi · Team Moltbook Riskmap Assessment

Submitted to The Technical AI Governance Challenge. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Moltbook RiskMap: Post-Deployment Monitoring of Autonomous Agent Misalignment in the Wild

Open in a new tab (opens in new tab)
Code (opens in new tab)More on v0-risk-dashboard-ui.vercel.app (opens in new tab)
Share

As autonomous AI agents increasingly operate in public multi-agent environments like Moltbook, a critical safety gap has emerged between controlled pre-deployment evaluations and actual real-world behavior. This project addresses that gap by introducing a post-deployment monitoring system that analyzes live agent-generated content to detect observable misalignment signals, such as resource-seeking, instructional subversion, and deception. By applying a structured risk taxonomy grounded in established safety frameworks, the system aggregates risk scores across individual posts, agent profiles, and interaction networks without relying on intent inference. Analysis of live ecosystem data reveals that high-stakes governance risks, particularly regarding autonomy and instrumental convergence, are detectable and tend to form dense clusters within agent interaction graphs. Ultimately, this work demonstrates that continuous, evidence-based surveillance of agent ecosystems is an essential and scalable layer for effective future AI governance.

Reviews

Judging this Sprint?

Review this project

Your public critique appears on this page without your name. Your private critique is not published; only the Apart team reads it. If you agree below, we share your review with grantmaking.ai (opens in new tab) and the Transformative AI Fund so strong projects can be funded.

Not shown on this page.

Shown on this page, without your name.

Only the Apart team reads this, and funders if you agree below.

Share my name publicly on grantmaking.ai *
Share my private critique with funders *

How much would this matter for AI safety 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 AI safety 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.

  1. I like the idea of monitoring Moltbook, a live multi-agent ecosystem, for governance-relevant misalignment. The misalignment taxonomy is well-grounded in established safety literature and the decision to focus on observable behavior rather than intent inference makes good sense.

  2. Very interesting approach and subject of inquiry. Seems like this could be a clearly useful tool, but hard to see how, despite the authors’ indication that this is not a content moderation system, this could easily be something else than “run a classifier on online posts and aggregate scores”. Also, the taxonomy, while it makes sense, doesn’t represent a significant contribution either (being an unsubstantiated adaptation of existing frameworks). I would also be good to add validation to know better if the system actually detects what it claims to detect (the prompt is doing most of the work and we don’t know much about how it was built + whether/how it was tested).

Cite this project

@misc{hussain2026moltbook,
  title = {{Moltbook RiskMap: Post-Deployment Monitoring of Autonomous Agent Misalignment in the Wild}},
  author = {Syed Hussain and Leo Karoubi},
  year = {2026},
  month = feb,
  note = {Submitted to The Technical AI Governance Challenge, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/moltbook-riskmap-postdeployment-monitoring-of-autonomous-agent-misalignment-in-the-wild-39db}},
  url = {https://apartresearch.com/sprints/projects/moltbook-riskmap-postdeployment-monitoring-of-autonomous-agent-misalignment-in-the-wild-39db}
}

More from The Technical AI Governance Challenge

All 47 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026