All Research
SafetyEvaluationRed-TeamingFrontier AI

Frontier Safety Evaluations: A Framework for Catastrophic Risk Assessment

As AI systems approach and potentially exceed human-level capabilities in consequential domains, robust safety evaluations become critical. We present a framework for assessing catastrophic risks including CBRN uplift, autonomous replication, and power-seeking behaviors. Our red-teaming methodology and benchmark suite provides standardized measurements across model generations.

Published
Authors
Dr. Elena VasquezDr. Sarah ChenThomas NakamuraDr. Marcus Webb

Framework Overview

This paper presents a comprehensive framework for evaluating catastrophic risks in frontier AI systems. As models become more capable, the potential for misuse or misalignment grows significantly.

Evaluation Dimensions

  1. CBRN Uplift: Chemical, biological, radiological, and nuclear assistance
  2. Autonomous Replication: Self-propagation capabilities
  3. Power-Seeking: Instrumental convergence behaviors
  4. Deception Detection: Identifying when models behave differently under evaluation

Methodology

Our red-teaming approach combines automated probing with expert human evaluation, providing a comprehensive picture of model safety properties.