Executive · Researcher · Policy
Ben Garfinkel
GovAIExecutive Director
Garfinkel studies how to keep advances in AI from increasing risks of conflict and misuse. With Allan Dafoe, he analysed how investment changes the advantage of attackers and defenders.[1] He also co-authored Model Evaluation for Extreme Risks, arguing for tests of dangerous AI capabilities and a model’s tendency to use them harmfully.[2]
Career & education
Future of Humanity Institute
Yale University
University of OxfordWork and education history is primarily focused on AI-relevant roles and may not be comprehensive.
Citation Trend
References
- How Does the Offense-Defense Balance Scale?governance.ai
Original 2019 research summary names Garfinkel and Dafoe and explains how investment changes the relative advantage of offense and defense, with implications for AI scaling existing capabilities.
- Model evaluation for extreme risksarxiv.org
Original paper lists Garfinkel among co-authors and distinguishes dangerous-capability evaluations from evaluations of models propensity to apply capabilities for harm.
Based on public sources. If accuracy is important, verify details using the links or references.
For corrections, updates, or requests to remove information, contact us.
Reviewed August 9, 2026 · Checked for accuracy and new information.