Skip to content
Apart Research
Apart Research
Contact us
Sprint projectSep 14, 2026Paris

Playing Through an AI Incident

Sofia Gallego · Team Nefinia

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

Read the report

Report: Playing Through an AI Incident

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

Little Agent Lab is a small browser game where you build a team of AI agents to complete a task while keeping a private item from leaking. Three short missions escalate from basic capability separation to a final one where two guard characters offer to watch the gate — only one actually does, and you find out by running the plan, not by reading a description. It's an early proof of concept aimed at giving people, especially kids, a playable feel for ideas like oversight and verification. Play it at sofiagallego.com/little-agent-lab.

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. Helping people understand why incidents occur - the concepts of monitoring, safeguards and permissions in practice and why they might fail - is important and could help make aspects of AI risks and incidents more accessible to new audiences.

    The paper could have a stronger link to potential impact by building on the idea it is communicating that 'a safeguard that is present but does not work' and extending this to when this might happen, particularly with respect to advanced AI systems which can find workarounds. So it could cover different failure modes - wrong setup, not actually applied, a system can subvert it intentionally.

    It would also have been interesting to see a rationale relating to the choice of aspects included for this initial prototype, as they relate to what the author considers most important to communicate to a non technical audience about AI incidents.

  2. This is a delightful idea that I'd love to see fleshed out more. As society grapples with helping children grow up with a technology that could theoretically destroy the world, I appreciate any project that attempts to make it easier to explain!

Cite this project

@misc{gallego2026playing,
  title = {{Playing Through an AI Incident}},
  author = {Sofia Gallego},
  year = {2026},
  month = sep,
  note = {Submitted to AI Incident Response Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/playing-through-an-ai-incident-d1bo}},
  url = {https://apartresearch.com/sprints/projects/playing-through-an-ai-incident-d1bo}
}

More from AI Incident Response Sprint

All 233 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026