How Blueprint Alliance Secures AI Agents Across Ecosystems

To support these real-world requirements, Alliance members commit to deep technical interoperability across open standards including MCP, Open Cybersecurity Schema Framework (OCSF), Shared Signals Framework (SSF) and Continuous Access Evaluation Profile (CAEP).

Okta expands identity governance for agents

To support the framework directly, Okta announced new capabilities within Okta for AI Agents to provide lifecycle management, runtime protection and instant containment.

“Moving fast with AI agents should not mean sacrificing control,” Ryan Barnes, Director of IT at MJS Packaging, says.

On the other hand, Ric Smith, President of Products and Technology at Okta, explains that access permissions make agents both powerful and dangerous.

He says: “For over a decade, Okta has continuously modernised how the workforce authenticates and connects to every app they use. That same discipline extends to AI agents.”

As part of this rollout, Okta introduced Shadow AI Agent Discovery for Endpoints to surface unmanaged local agents running on employee devices.

The provider also expanded connection governance through Agent SSO, Agent-to-Agent Connections rules, visual Configuration Designer tools and Resource Access Certifications.

Runtime visibility is further enhanced via the Agent Gateway, which enforces policies directly within active execution paths.

Looking ahead, Okta plans to expand Kill Switch capabilities to the gateway, allowing administrators to revoke active tokens and terminate rogue sessions instantly.