Securing data centers in the agentic AI era
Source Tenable Research
Description
This analysis addresses the evolving security challenges faced by data centers in the era of agentic AI, where AI systems autonomously execute business decisions. Data centers have become critical national infrastructure, integrating legacy operational technology (OT) and extended IoT (xIoT) devices that were not originally designed with modern security in mind. The threat landscape includes AI-powered attacks targeting both digital assets and physical infrastructure, such as cooling and power systems, potentially causing operational denial and costly downtime. Attackers may exploit vulnerabilities in third-party vendor portals or engineering workstations to pivot into process control layers, disrupting physical systems. The complexity of securing data centers is heightened by the need to maintain near-perfect uptime while managing rapid patch cycles and protecting identity-based network communications. Regulatory frameworks increasingly require continuous, real-time monitoring and proactive exposure management across IT, OT, identity, and AI domains. A comprehensive security strategy involves asset discovery, attack path anticipation, risk prioritization, identity perimeter protection, third-party risk monitoring, and compliance streamlining.