Driving thought leadership to shape responsible and effective deployment and oversight.
Convening chief science officers, technical leaders, and AI system developers to build consensus around the safe and effective deployment of advanced AI systems — including world models, multi-modal models and AI agents. It focuses on fostering shared technical standards, advancing safety protocols, and shaping the design of AI systems that are robust, aligned, and adaptable to complex real-world contexts.

Examine the current landscape and future potential of AI agents, focusing on their capabilities, readiness levels, and autonomy.

Build consensus on the emerging frontier of generative AI systems, from large language models to large multi-modal models, and how they power a new era of AI agents.

Address technical governance challenges associated with the integration of AI agents into daily life and into various sectors and industries.
As organizations scale AI agents, they face growing challenges in defining and enforcing clear authorization, accountability, and oversight across increasingly autonomous systems. This report introduces the Agent Capability and Authorization Profile (ACAP), a governance framework that makes AI agent actions auditable, enforceable, and aligned with human responsibility throughout the lifecycle. It provides a practical roadmap to support trusted AI adoption, scalable deployment, and standardized governance of agentic systems.
AI Global Alliance
A World Economic Forum Initiative