Managing the Explosion of Non-Human Identities
Combined Session
Tuesday, October 06, 2026 10:30—11:30
Location: Ballroom
Log in to download presentations
Tuesday, October 06, 2026 10:30—11:30
Location: Ballroom
Enterprises now manage identities that span employees, partners, customers, workloads, and emerging AI agents, each with different risk profiles and lifecycle demands. Disconnected governance, access, and threat-detection tools make it difficult to enforce policy consistently or respond to attacks quickly. An identity security fabric provides a unified architectural layer that centralizes visibility, standardizes controls, and automates responses across every identity and system.
Stephen will demonstrate how to operationalize an identity security fabric using Okta’s platform, showcasing orchestration-driven response automation and protection before, during, and after authentication. He will address securing AI agents, reduce misconfigurations, and achieving unified visibility across cloud, on-premises, and hybrid ecosystems.
This session sets a strategy for governing the identities that already outnumber people - service accounts, API keys, tokens, secrets, and now autonomous AI agents that authenticate, reason, spawn, and act with standing privilege and no clear owner.
It characterizes non-human identity (NHI) and agentic AI as a distinct class of privileged risk, applies a maturity model to locate where an organization actually stands, and prioritizes the near-term controls that establish a defensible baseline.
It situates this against the regulatory reach of NIS2 and DORA, and against the oldest and most exposed non-human population of all — operational technology (OT), where shared service and vendor accounts still run on static credentials unchanged since installation.
Machine identities have become the dominant identity type in modern enterprises, yet many organizations continue to manage them with human-centric approaches. This panel explores how enterprises can adapt their identity strategies to manage machines, workloads, services, and AI systems at scale while maintaining visibility, ownership, and control.
