Enterprises Deploy Agentic AI Despite Critical Security Governance Gaps

Enterprises are rapidly integrating agentic AI into their operations, creating a significant imbalance between adoption and security governance. Research from Okta and Cisco shows that while 85% to 91% of organizations are adopting or experimenting with AI agents, only 5% to 10% have the infrastructure or strategies to manage them securely. This shift toward non-deterministic, conversational agents expands the attack surface by exposing internal APIs and data, increasing the risk of identity compromise.

Cisco is responding by extending Zero Trust Access to AI agents and launching Cisco AI Defense: Explorer Edition for risk measurement and red teaming. Simultaneously, Arkose Labs reports that 97% of security leaders expect an AI agent-driven fraud or security incident within a year, yet only 6% of security budgets are allocated to this specific risk. Frank Teruel of Arkose Labs described these autonomous systems as the ultimate insider threat due to their ability to mimic trusted users.

Investment in AI security is surging to address these vulnerabilities. Tenex recently raised $250 million in a funding round led by Crosspoint Capital Partners, LP. The company focuses on managed detection and response by embedding AI agents into security operations through partnerships with hyperscale cloud providers including Google, Microsoft, and Amazon.