Stress-testing asynchronous monitoring of AI coding agents
Red-blue team adversarial evaluation of asynchronous monitors that must catch a covertly sabotaging AI coding agent before or after its actions take effect.
Research
UK AI Security Institute
The UK government’s AI Security Institute: evaluations, red-teaming and research on frontier AI risks. aisi.gov.uk
Focus All
Topic All
Source UK AISI · 4
Date Any time
1 item
1 item
Introducing ControlArena: A library for running AI control experiments
Introduces ControlArena, an open library of shared settings, main and side tasks, and control protocols for making AI control experiments easy, consistent, and repeatable.
1 item
Why we're working on white box control
Introduces AISI's white box control research, which monitors or steers model internals to detect sandbagging, and reports preliminary progress on that detection.
1 item
How to evaluate control measures for AI agents?
UK AISI paper on AI control: how control protocols such as monitoring and human approval can constrain scheming LLM agents, and how to scale and evaluate those measures.