Cybersecurity Recommendations for Securing AI Agents

This paper provides a set of controls, a secure-by-default checklist, and a data classification and handling guide that can serve as important inputs when securing agentic AI.

Published: 2026-09-14