How Anthropic uses Claude in Cybersecurity

Anthropic’s cybersecurity team is leveraging Claude Code to build CLUE, a custom detection and response platform designed to overcome the inefficiencies of traditional security investigations. By integrating directly with internal data warehouses, Slack messages, and codebases, CLUE automates the complex process of jumping between multiple tools and query languages. The platform can autonomously generate investigation plans, execute queries across diverse data sources, and identify malicious activity—such as a simulated privilege escalation from a Russian data center—while suggesting security posture improvements. This AI-driven approach has significantly accelerated internal development, enabling a new hire to build a suppression engine in one week rather than two months. This shift allows security practitioners to move beyond manual data processing toward a more research-oriented role, utilizing AI to gain unprecedented visibility into complex systems and manage immense data volumes with greater autonomy.