Agentic AI & Autonomous Systems

About this course

This course provides a rigorous, non-technical understanding of agentic and autonomous AI systems for professionals who evaluate, govern, approve, or deploy AI-driven workflows in real organizational environments.

Participants learn what makes a system agentic, how autonomy emerges from the combination of models, planning, memory, tools, and feedback, and why agentic systems introduce qualitatively new risks compared to predictive models or standalone foundation models. The course explains how modern agentic systems operate over time, take actions in the world, coordinate with other systems, and fail in ways that are often invisible until damage occurs.

Rather than treating agents as “LLMs with tools,” the course builds durable mental models grounded in decision theory, planning, and control. Participants examine autonomy boundaries, delegation risks, escalation failures, and enterprise-level breakdowns such as runaway execution, silent objective drift, and loss of human accountability.

The focus is not on building agents or using frameworks, but on understanding how autonomous behavior emerges, why control is fragile, and how organizations must design limits, oversight, and governance before deployment.

This course is designed for leaders, legal and compliance professionals, product owners, risk managers, and business teams responsible for approving or supervising autonomous AI systems, without writing code or understanding mathematics.