RSA, a cybersecurity company focused on identity and access management, launched an AI agent identity platform, RSA Agent ID, on October 7 at World Summit AI in the Netherlands.

The platform intends to help regulated organizations discover, register and govern AI agents throughout their lifecycles. The platform is designed to assign each agent a named human owner, risk tier and lifecycle state while tracking the authority behind actions.

RSA Agent ID includes tools to: 

• discover known and unknown AI agents and Model Context Protocol (MCP) servers
• enforce policies on agent actions 
• maintain auditable records of governed activity

Its AI/MCP Gateway can run in RSA’s cloud or inside an organization’s own boundary, supporting cloud, hybrid or on‑premises deployments.

AI Agents Become Governable Enterprise Actors

The launch reflects a growing AI governance issue as organizations give agents access to business systems and data. 

Unlike conventional tools, AI agents can make decisions and take actions based on instructions and the information available to them. For chief data officers (CDOs), that creates a need to connect governance with existing controls over data access. 

Knowing which agents operate in an environment, who owns them and what information they can access can become part of the organization’s broader AI governance framework.

RSA’s approach of high‑risk actions requiring a named, authenticated human, and governed actions being recorded and linked to the person authorizing them is commendable as it gives CDOs a clear model for accountability.  

It ensures every consequential agent action can be traced to an accountable owner.  

It also makes it easier to justify those actions against data access policies and regulatory requirements.