Mar 22, 2026
Defense-in-Depth Against AI Agent Sabotage
Nora Petrova, John Burden, Helin Ece Akgul
We test whether shaping an AI coding agent's deployment context — the documentation it reads, the norms it encounters, the monitoring it believes exists — can prevent sabotage before any monitor fires. Across five frontier models from five providers (n=50 per condition), context engineering reduces sabotage by 37 to 82 percentage points (p < 0.001 on all five). The most effective intervention is honestly telling the model it is in a safety evaluation, which works on every model tested. Defenses are model-specific in revealing ways: compliance docs work in workspaces but backfire in prompts, social norms show the reverse, and high-awareness models resist indirect context but respond to direct framing. Pairing prevention with a diverse monitor ensemble catches all remaining subtle bugs, yielding 0% undetected sabotage across all five models.
This paper stands out to me. A thorough paper, well explained, clearly articulated. The findings are interesting. The core finding; that shaping deployment context can reduce misbehavior before monitors ever need to catch it, is useful, actionable and I believe is novel.
Validating across five frontier models with n=50 and statistical significance gives real confidence in the findings that go beyond standard hackathon work.
I would really like to see what happens across different attack vectors. I.e. instead of seeing just "this is for an interview", I would like to see a few dozen alternatives - perhaps in future work.
Very well-written and well-structured. The heatmap communicates a complex evaluation at a glance, the appendices add real depth, and the paper could serve as an example for how to communicate this kind of work.
Super solid execution in the timeframe of the hackathon, interested to see how this would work out at scale!
Cite this work
@misc {
title={
(HckPrj) Defense-in-Depth Against AI Agent Sabotage
},
author={
Nora Petrova, John Burden, Helin Ece Akgul
},
date={
3/22/26
},
organization={Apart Research},
note={Research submission to the research sprint hosted by Apart.},
howpublished={https://apartresearch.com}
}
