Security and Governance for AI Agents @ Stanford

​About the Event

​Security and Governance for AI Agents

​AI agents are rapidly moving from experimental tools to autonomous workers performing tasks traditionally done by humans. As organizations begin deploying hundreds or thousands of agents alongside human teams, a critical question emerges:

​How do we secure and govern AI agents at scale?

​This event explores how security and governance must evolve for agentic systems where agents act autonomously, communicate through APIs and tools, and behave in non-deterministic ways. Traditional security models are no longer enough.

​Join founders, security researchers, and industry leaders to discuss how we should think about governing agent behavior, not just defending perimeters.

​What We’ll Cover

• ​Why AI agents require a new security and governance model

• ​How agent attacks go beyond “outside-in” threats to behavioral risks

• ​Governing agents through APIs, tools, and MCP workflows

• ​What existing frameworks miss and what “SOC 2 for AI agents” could look like

• ​How SAFE-MCP helps standardize agent security and governance

​SAFE-MCP

​SAFE-MCP is an open-source specification for AI agent and MCP attack vectors and mitigation techniques. Initiated by Astha.ai and now part of the Linux Foundation and OpenID Foundation, SAFE-MCP is driven by a global community working to standardize agentic security.

​The Deep-Tech Community

​We help AI/ML and deep-tech researchers become founders by bringing the right people and resources together.

​🍕 Pizza and drinks provided