Okta introduces AI agent runtime gateway for enhanced identity security

As reported by Silicon Angle, Okta Inc. has launched an AI agent runtime gateway, a new component of its identity security strategy that moves beyond initial access control to govern the actions of autonomous AI agents within an organization. This gateway sits in the path of attempted actions, enabling real-time enforcement of authorization decisions.

Ric Smith, president of products and technology at Okta, explained that this move signifies a significant shift for the company, allowing them to be inline and make more authorization decisions. The gateway's inline position enables threat monitoring in addition to permission granting, a capability bolstered by Okta's acquisition of Permiso Security. To achieve full visibility into AI agent behavior, the system captures both the initial prompt and the proposed actions.

Okta plans to integrate intent-based authorization into the gateway in 2027, using an agent's context and permitted tools to inform access decisions. This approach aims to address the complex challenge of narrowly scoping permissions without hindering legitimate operations, a key area of ongoing research for Okta as they expand their cybersecurity offerings.

Source: Silicon Angle