Locking Down the Machine: Securing and Accelerating Enterprise AI Agents with Ory

Enterprises are deploying autonomous AI agents and non-human identities (NHIs) at a pace that has outrun traditional security controls, and leadership isn't asking security teams to slow that down. They're asking CISOs to make it safe without making it stop. There are various challenges enterprises face today that tie into agent IAM and security. For example, static API keys leak into public repositories, agents execute on unauthorized commands – this goes to show that traditional defenses were never built for machine-to-machine trust at an agentic scale. Join Ory for an exclusive deep dive into the next generation of identity security. In this session, we will explore Ory’s newly launched, cutting-edge Agent IAM portfolio—including Ory Talos, Ory Agent Security, and Ory Agent DX—alongside our strategic integration with Tetrate. Learn how to replace vulnerable, static credentials with programmable, short-lived tokens, and discover how to embed absolute governance right at the point where an AI agent decides to act. What You Will Learn: - Dynamic Credentialing with Ory Talos: Not every workload carries the same risk — and your credentialing strategy shouldn't treat them the same. Ory Talos gives you a range of token formats, from JWTs to Macaroon-based tokens, each scoped to least privilege and issued dynamically. Match your credential strategy to your risk posture without overhauling what's already working. - Securing the Action with Ory Agent Security: A look at our patent-pending architecture that embeds a security control plane directly into the agent harness, allowing you to intercept, evaluate, and audit commands before they execute. - Accelerating Secure Development with Ory Agent DX: How free developer plugins bring Ory’s identity infrastructure directly into AI coding tools (like Claude Code and Gemini CLI), allowing developers to scaffold secure authentication natively using natural language. - Containing the Blast Radius with Ory + Tetrate: Agents need broad access to be useful. That's exactly what makes them dangerous. Ory locks down agent identity and least-privilege permissions at the application layer; Tetrate enforces them at the network gateway on every live call. High-risk actions are intercepted before they execute, require human approval, and elevated access expires after a single use — so you can say yes to AI deployment without handing agents unlimited access.