Contributors Acknowledgements Animashree (Anima) Anandkumar Bren Professor of Computing and Mathematical Sciences, California Institute of Technology (Caltech) Mandanna Appanderanda Nanaiah Head, Infosys Responsible AI, North America, Infosys Nebahat Arslan Director, Group General Counsel and Partnership Officer, Women in AI Mennatallah El-Assady Professor of Interactive Visualization and Intelligence Augmentation, ETH Zurich Ricardo Baeza-Yates WASP Professor, KTH Royal Institute of Technology, Sweden Amir Banifatemi Chief Responsible AI Officer, Cognizant William Bartholomew Director of Public Policy, Responsible AI, Microsoft Aaron Bawcom Field Chief Technology Officer, Invisible Technologies Lisa Bechtold Group Head of Governance, Zurich Insurance Pete Bernard Chief Executive Officer, EDGE AI FOUNDATIONFabio Casati Lead, AI Trust and Governance Lab, ServiceNow Kevin Chung Chief Strategy Officer, Writer Cathy Cobey Global Trusted AI Advisory Leader, EY Ben Colman Co-Founder and Chief Executive Officer, Reality Defender Sakyasingha Dasgupta Founder and Chief Executive Officer, EdgeCortix Umeshwar Dayal Senior Fellow and Senior Vice-President, Hitachi America; Corporate Chief Scientist, Hitachi Mona Diab Director, Language Technologies Institute, Carnegie Mellon University Yawen Duan AI Safety Research Manager, Concordia AI Gilles Fayad Adviser, Institute of Electrical and Electronics Engineers (IEEE) Claudia Fischer Public Policy Planning, Global Affairs, OpenAIThe World Economic Forum’s AI Governance Alliance Safe Systems and Technologies working group convenes chief science officers and AI producers to advance thought leadership surrounding AI agents, from their architecture to applications, social implications, guardrails and governance structures. This initiative promotes the development of safety mechanisms and encourages collaboration on best practices for the design and implementation of AI systems. World Economic Forum Benjamin Cedric Larsen Initiatives Lead, AI Safety, Centre for AI ExcellenceCapgemini Olivier Denti Data Architect, AI, Capgemini Invent Jason DePerro Human-AI Collaboration Director, Capgemini Invent Jeanne Heuré Vice-President, Digital Trust & Security, Capgemini Invent Raymond Millward GenAI for R&D Technical Solution Lead, Capgemini Engineering Efi Raili Safety Authority, Technology and Innovation, Capgemini Engineering AI Agents in Action: Foundations for Evaluation and Governance 30

AI Agents in Action Foundations for Evaluation and Governance 2025