AI agent security is top customer concern, Okta exec says

Okta is seeing an overwhelming surge in customer demand for securing AI agents, which the company's President and COO Eric Kelleher now identifies as the single biggest identity-related security concern for enterprises, reports CRN.

In an interview, Kelleher stated that management and security of agentic technologies is "consuming our conversations with customers right now," and that clients view it as their "most important strategic need" around identity today. This urgency is underscored by Okta's own survey data, which found that while 90% of organizations have deployed AI agents, only 10% are confident those agents are properly governed and secured.

To address this gap, Okta has elevated agentic identity to a "first-class citizen" in its platform, extending its existing identity governance, credential vaulting, and policy enforcement capabilities to cover non-human entities. This allows customers to provision and de-provision agents in a just-in-time manner, eliminating standing permissions and applying the same security policies as for human users.