Skip to content
Apart Research
Apart Research
Contact us
Sprint projectJan 11, 2026Barcelona

AI Swarms Manipulation: How Coordinated Infiltrator Agents Shift Community Beliefs

Publius Dirac, Anantha Shakthi Ganeshan Thevar, Babita Singh · Team AI Swarms Manipulation

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

Read the report

Report: AI Swarms Manipulation: How Coordinated Infiltrator Agents Shift Community Beliefs

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

We simulate how a small group of sophisticated AI agents ("infiltrators") can manipulate a larger community of less capable AI agents into adopting a specific belief. It models real-world information influence campaigns to help understand vulnerabilities to coordinated manipulation. We find that even a single infiltrator achieves high belief adoption, but pre-seeded dissenters act as "antibodies" that can reverse adoption over time, suggesting viewpoint diversity provides natural resistance to manipulation.

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 would like to see this project continue at a larger scale as we see more agent-to-agent interactions on the horizon. And perhaps we should see this project continue on actual social networks with interactions with humans. It feels like an important area, and the result of anti-bodies being able to regularize manipulation seems like something we should be exploring more. Great job.

    Love to see a verification test on a theoretical work. As far as I can tell, this is the first attempt at seeing how agents can manipulate each other in social platform dynamics. The anti-bodies is a great exploration and result.

    The execution is solid considering the time constraints, but there is clear areas of improvement here. Would have boosted the score if simulation ran with larger populations and multiple beliefs were tested. Would also have been impactful to do multiple runs to see reproducibility or statistical significance.

    Well explained. Good structure of report. Useful graphics. Clear in the limitations. Clear in the approach.

    Read full reviewShow less

  2. Studying coordinated influence and manipulation in multi-agent systems is highly relevant given the current trajectory of LLM development and deployment. It's interesting that a few "antibodies" can partially negate a coordinated attack, and I'd like to see this explored with higher sample sizes and repeated runs. I think the biggest weakness is clearing agents' context each timestep (which the authors acknowledge due to compute constraints) -- this makes the setup meaningfully different from real dynamics.

Cite this project

@misc{dirac2026ai,
  title = {{AI Swarms Manipulation: How Coordinated Infiltrator Agents Shift Community Beliefs}},
  author = {Publius Dirac and Anantha Shakthi Ganeshan Thevar and Babita Singh},
  year = {2026},
  month = jan,
  note = {Submitted to AI Manipulation Hackathon, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/ai-swarms-manipulation-how-coordinated-infiltrator-agents-shift-community-beliefs-9ek9}},
  url = {https://apartresearch.com/sprints/projects/ai-swarms-manipulation-how-coordinated-infiltrator-agents-shift-community-beliefs-9ek9}
}

More from AI Manipulation Hackathon

All 70 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026