The U.S. Cybersecurity & Infrastructure Security Agency (CISA), in collaboration with the Australian Signals Directorate’s Australian Cyber Security Centre and “other international and U.S. partners,” has released guidance for organizations on adopting agentic artificial intelligence (AI) systems.
The guidance, titled “Careful Adoption of Agentic AI Services,” outlines “key security challenges and risks associated with agentic AI,” “provides actionable steps for designing, deploying, and operating these systems safely,” and “helps organizations align AI risk management with existing cybersecurity frameworks and strengthen oversight as agentic AI adoption grows,” according to CISA.
Read the full guidance document here.
