Poster · Sydney
Attack Selection In Agentic AI Control Evaluations Meaningfully Decreases Safety
- When
- Thu 10 Dec 2026, 10:00–13:00 AEDT
- Where
- Hall 1-4, ICC Sydney
Exhibition Centre, Level 1 - Session
- Sydney Poster Session 5
- Authors
- Catherine Ge-Wang, Tyler Crosse, Benjamin Hadad, Joachim Schaeffer, Ram Potham, Tyler Tracy
- Organisations
- University of Oxford, Georgia Institute of Technology, University of North Carolina at Chapel Hill, MATS, Redwood Research
- QRS classification
- Agentic AI, Safety, fairness and privacy