Thursday, October 22, 2026
9:00 AM – 5:00 PM EDT
Advancing the Science of AI Agent Evaluation
About this event
A two-day research workshop on shared measurement methods, tools, standards, and governance frameworks for evaluating increasingly autonomous AI agents.
Microsoft Research and Carnegie Mellon University's AI Measurement Science & Engineering Center convene researchers, industry practitioners, civil-society participants, and government representatives at the Microsoft Research New York City Lab. Keynotes, lightning talks, panels, and working groups examine general and domain-specific agent evaluation, tooling, accountability, and governance. The workshop is designed to produce public syntheses of open problems and emerging practices.
Organized by
Microsoft's research organization, convening researchers and practitioners around artificial intelligence, computing, evaluation, governance, and responsible technology.