Okta's revenue growth highlights a new security challenge: companies must now manage and secure not only employee accounts but also a rapidly expanding population of autonomous AI agents with access to sensitive data.
As artificial intelligence becomes integrated into business operations, organizations encounter new security requirements. Autonomous AI agents-software capable of reading emails, updating records, and executing business processes-now function as digital workers, necessitating identity controls and oversight comparable to those for human employees.
Okta has introduced Agent SSO as a separate subscription, leveraging Cross App Access (XAA) to treat AI agents as a distinct class of digital identities requiring dedicated access controls.
AI Agents: Expanding the Attack Surface
AI agents operate continuously and can perform complex tasks without direct oversight, increasing both efficiency and risk. If an AI agent's credentials are compromised, attackers may gain broad system access. A 2023 CyberArk study found that 79% of organizations expect machine identities to increase in the coming year, with many anticipating growth of 50% or more. Each AI agent requires unique authentication, permissions, and audit trails, making identity management more complex.
Okta for AI Agents includes features such as agent registration in a central directory, lifecycle management, and deactivation capabilities. The platform supports onboarding, access certification, approval workflows, and short-lived token issuance, providing a full IAM control cycle for both sanctioned and unsanctioned AI agents deployed across the enterprise.
Automation Raises Security Stakes
Security incidents involving AI agents are an increasing concern. Verizon's 2026 breach report noted that shadow AI-AI systems deployed without formal oversight-rose to 45%, heightening data loss risks. Software vulnerabilities accounted for 31% of initial breach access, and automation enables attackers to move rapidly within networks. IBM reported the average global cost of a data breach now exceeds $5 million, with attacks on AI models sometimes resulting in even greater losses.
Okta's financials reflect the urgency of these risks. The company generated $234 million in operating cash flow and $227 million in free cash flow for the quarter. Current remaining performance obligations increased 14% to $2.585 billion. Okta projects full-year revenue between $3.22 billion and $3.23 billion, underscoring the scale of the identity management market as AI adoption accelerates.
Identity Management for Non-Human Actors
Identity management systems underpin daily business operations, verifying logins and enforcing security policies. As AI agents proliferate, organizations must consider not only "Who has access?" but also "What has access?" Misconfigured permissions for AI agents-such as access to customer payment or health data-can result in significant breaches.
Investor interest has grown: Okta shares rose after its latest results and guidance exceeded expectations, according to The Wall Street Journal. However, the market for AI agent security remains nascent. Some firms may delay adopting these controls, and large software vendors could integrate identity features directly into their platforms. Okta is positioning itself to provide identity solutions for both human and AI digital workers.
Financial pressures add complexity. With U.S. national debt surpassing $40 trillion and interest costs rising, as detailed in this analysis, companies must balance innovation, risk management, and cost control.
Identity management is now central to cybersecurity strategy. It involves verifying users and systems, assigning permissions, and monitoring activity to prevent unauthorized access or data leaks. As AI agents become more prevalent, organizations must update technical controls, policies, and audit practices to address non-human actors. The effectiveness of these measures will increasingly determine how well sensitive information is protected in an era of rapid automation.
