AI Safety Week | Day 2: From Evaluation to Evidence
Tuesday, November 10, 2026 · 9am – 10.30am
Session
AI Safety Week | Day 2: From Evaluation to Evidence
Tuesday, November 10, 2026
9am – 10.30am · Eastern
Working session
Delegate registration is being released in phases. Eligible delegates will receive access instructions when registration opens to them.
Information
Led by METR with support from Rethink Priorities and Center for Mind, Ethics, and Policy, participants get an inside look at how AI capability evaluations are designed, conducted and interpreted. The session will examine how researchers identify the right questions, develop methodologies and navigate uncertainty in their findings. We'll also consider what policymakers, funders and researchers need from evaluations to make informed decisions, and where important questions remain unanswered.
Convened with
METR
Independent research nonprofit developing scientific methods to evaluate the capabilities and risks of advanced AI systems. Its empirical assessments help inform decisions about AI development, deployment and oversight.
Rethink Priorities
Research-driven think-and-do tank identifying effective solutions to pressing global challenges. Its work spans AI, global catastrophic risks, animal welfare and global development, helping decision-makers direct resources towards meaningful impact.
Center for Minds, Ethics, and Policy
Advances interdisciplinary research on the consciousness, sentience and moral status of nonhuman minds, including animals and AI, to inform ethical understanding and policy decisions.