CrowdStrike Warns AI Adoption Is Creating ‘Underdefended’ Attack Surfaces

CrowdStrike’s 2026 Threat Hunting Report warns that attackers are using AI to accelerate discovery, exploit development, and operational scale — shrinking the window between vulnerability disclosure and exploitation to hours. Adversaries increasingly abuse legitimate identity and cloud services, target software package registries and AI frameworks (including compromises of >130 AI packages), and run high-volume automated attacks, forcing organizations to move beyond traditional patch cycles and perimeter defenses.