Ory Agent Security Brings Identity Enforcement Into the Agent Harness

Customers deploy Ory Agent Security because, “extending trust to non-human identity is vital”

Ory, one of the world’s most widely adopted platforms for customer, B2B, and agent identity management, announced the general availability of Ory Agent Security, the first Agent IAM control plane designed to secure AI agents at the point where they take action.

Partnering with Ory allows us to deliver real, deterministic agent security today, not just future roadmap promises.”

— John Poptsis, Managing Director at TeleGlobal

Existing AI security controls can isolate an agent, monitor its activity or control traffic passing through a gateway, but they generally sit outside the agent itself. That leaves a fundamental gap between knowing what an agent did and controlling what it is allowed to do. Controlling what they are allowed to do is critical as AI agents are moving into enterprise environments faster than organizations can govern them.

Also Read: IT Tech Pulse Exclusive Interview with Ken Claffey Chief Executive Officer and President of VDURA

Ory Agent Security closes that gap by enforcing identity and access policies directly inside the agent harness. It evaluates a tool action against policy before execution, allows or denies the action, records that decision, and binds actions to an accountable owner and delegated agent identity. This creates a chain of accountability from the human who owns an agent through the agent, any sub-agents and the tools they invoke.

“AI agents act with human credentials and leave a record that cannot answer the only question that matters. If an AI agent took a destructive action in your environment tomorrow, what would your audit log say, and could you name the human who set it running?,” said Jeff Kukowski, CEO of Ory. “The security model has to evolve accordingly. Ory Agent Security puts control where the action happens, so organizations can enforce what their agents are allowed to do, and are actually doing, rather than simply observing it afterward. This is the difference between an agent program you can govern and one you are hoping stays contained.”

Ory Agent Security supports 11 leading agent harnesses and 14 agent SDK frameworks, including Claude Code, LangChain/LangGraph, OpenAI Agents SDK and Vercel AI SDK. This allows teams to apply the same identity and authorization model across the different agents and frameworks they use including custom-built agents.

Ory believes the best place to start getting control of agents is to “turn the lights on” so organizations can see and understand what agents are doing. That’s why the platform starts in observe mode by design, giving teams full visibility into the humans, agents, sub-agents and tools operating in their environment without blocking developers or slowing existing workflows. Organizations move into enforcement on their own timeline, applying controls that range from broad guardrails to fine-grained permissions. Policies can be revoked immediately, and every decision is captured through OpenTelemetry to support audit and security workflows.

Also Read: IT Tech Pulse Exclusive Interview with Michael Jack Chief Revenue Officer and Co-Founder of Datadobi

Ory Agent Security is also designed for organizations that cannot put agent activity into a black-box security service. It can be self-managed in cloud, on-premises or air-gapped environments, or deployed through the Ory Network. The platform is built on standards including OAuth 2.1, OIDC, Zanzibar-style permissions and OpenTelemetry, allowing it to work alongside existing identity, security and infrastructure investments.

This approach builds on Ory’s identity infrastructure, which supports more than 3.25 billion identities, 4.4 billion transactions per day and more than 34,000 production deployments.

“Our clients trust TeleGlobal to manage and secure their critical technology infrastructure securely and with confidence, and as AI agents with non-human identities become embedded across the enterprise, identity, authorization, and governance become fundamental to maintaining that trust. Ory is a critical component of the TeleGlobal Compass™ framework because it provides enterprise-grade identity and access controls required to extend security and governance within the agentic AI environment. By integrating Ory into our managed services, we help clients move from AI experimentation to secure, governed AI adoption at scale, while preserving the flexibility to choose the AI platforms, agent frameworks that best support their business. Partnering with Ory allows us to deliver real, deterministic agent security today, not just future roadmap promises. More importantly, because Ory supports a wide range of agent harnesses and SDK’s, we can give our customers complete flexibility and choice in how they build, without forcing them to compromise on control or governance.” said John Poptsis, Managing Director at TeleGlobal.

Write to us [⁠wasim.a@demandmediaagency.com] to learn more about our exclusive editorial packages and programmes.