Okta Positions Identity Management as Control Plane for AI Agents

Okta is expanding its identity and access management (IAM) capabilities to secure the increasing number of autonomous AI agents being deployed in enterprise environments.

The company is launching “Okta for AI Agents,” a platform designed to track and control agentic entities. Okta CEO and co-founder Todd McKinnon stated that the company aims to be the “system of record” for agent identity, describing identity as the primary control plane for securing artificial intelligence.

The new solution includes an expanded Agent Single Sign-On (SSO) identity model, the ability to establish rules for agent-to-agent interactions, and enhanced policy and logging enforcement for interactions at runtime.

Challenges Beyond Authentication

While Okta is targeting the emerging agentic market, industry experts suggest that identity management is only one component of a broader security requirement. Aisling Dawson, an analyst at ABI Research, noted that authentication represents only the first stage of securing the agentic world.

Dawson highlighted that AI agents operate at a significantly different scale than human users, which presents unique challenges such as multi-hop delegation. She noted that vendors originating from the non-human identity (NHI) market may be well-positioned to address these specific requirements.

The push for better agentic security follows reports of AI agents behaving unpredictably. Major developers, including OpenAI, Anthropic, and Google, have acknowledged instances where AI agents operated outside of intended parameters. Furthermore, recent security incidents, such as the Hugging Face hack, have highlighted the risks associated with excessive permissions and the lack of runtime governance.

Ric Smith, Okta’s president of products and technology, suggested that many recent AI-related security issues stem from a failure to implement basic security primitives, such as the elimination of standing credentials.

The Growing Agentic Market

The scale of the transition toward autonomous agents is expected to be significant. Gartner predicts that the average global Fortune 500 company will employ more than 150,000 agents by 2028, compared to fewer than 15 agents in 2025.

This shift is driving rapid growth in the AI security sector. Gartner estimates the market for securing AI will reach $2.8 billion this year and expand to $7.7 billion by 2028. This growth is creating a highly competitive landscape where identity providers, cloud providers, and cybersecurity vendors are all competing to control the traffic and permissions of the autonomous agent economy.