In today’s cybersecurity landscape, a staggering 88% of organizations have reported AI agent security incidents, yet only 22% treat these agents as independent identities, creating a significant security gap. Ariel Kadyshevitch from Okta highlighted this issue at the Sydney Identity, Authentication and Access Management Summit 2026, emphasizing the urgent need for enhanced visibility and strategic governance of non-human identities. As enterprises embrace AI, identity has evolved into a crucial security perimeter, shifting from IT tool to core business strategy. The proliferation of “shadow AI” poses risks to sensitive data, necessitating proactive measures. Okta’s recent innovations, including a universal logout for AI agents and partnerships for cross-platform identity governance, aim to mitigate these risks. Kadyshevitch advocates for industry standards and emphasizes starting with visibility, lifecycle management, and governance to secure enterprise environments effectively. In essence, organizations must prioritize understanding and managing their growing inventory of non-human identities to prevent security breaches.
Source link
