AI Adoption Creates Massive Security Blind Spot for Enterprise Leaders

The report, which surveyed 200 technology and security leaders, reveals a troubling disconnect between rapid AI implementation and the infrastructure required to defend it. While 78% of respondents face significant pressure to ship AI capabilities, only 15% express high confidence in their current security stack. The core of the problem lies in the shift from human-centric to machine-centric identity.

Leaders identified non-human identities—such as AI agents, models, and API interactions—as their most significant vulnerability. This aligns with broader market trends: Gartner’s 2025 Machine Identity survey found that 42% of organizations lack a formal strategy for managing machine identities, despite a surge in their use. NetFoundry CEO Galeal Zino notes that the industry is still relying on legacy tools like VPNs and firewalls, which were never built for the automated, machine-driven traffic patterns defining modern enterprise environments. Consequently, 85% of survey participants are now actively seeking new strategies to secure these non-human workloads, as they expect the total attack surface to grow by another 14% over the coming year.