Optimizing AI Agent Attacks With Synthetic Data

MATS Fellow:

Jonathan Kutasov, Chloe Loughridge, Tyler Tracy

Authors:

Chloe Loughridge, Paul Colognese, Avery Griffin, Tyler Tracy, Jon Kutasov, Joe Benton

Citations

0 Citations

Abstract:

As AI deployments become more complex and high-stakes, it becomes increasingly important to be able to estimate their risk. AI control is one framework for doing so. However, good control evaluations require eliciting strong attack policies. This can be challenging in complex agentic environments where compute constraints leave us data-poor. In this work, we show how to optimize attack policies in SHADE-Arena, a dataset of diverse realistic control environments. We do this by decomposing attack capability into five constituent skills -- suspicion modeling, attack selection, plan synthesis, execution and subtlety -- and optimizing each component individually. To get around the constraint of limited data, we develop a probabilistic model of attack dynamics, optimize our attack hyperparameters using this simulation, and then show that the results transfer to SHADE-Arena. This results in a substantial improvement in attack strength, reducing safety score from a baseline of 0.87 to 0.41 using our scaffold.

Recent research

What Happens When Superhuman AIs Compete for Control?

Authors:

Steven Veld

Date:

January 11, 2026

Citations:

0

AI Futures Model: Timelines & Takeoff

Authors:

Brendan Halstead, Alex Kastner

Date:

December 30, 2025

Citations:

0

Frequently asked questions

What is the MATS Program?
How long does the program last?