AIDU-AGENT-104
Delivery Type: Live, instructor-led Remote or In person
Prerequisite: Foundation Models, LLMs & Multimodal AI
This course provides professionals with a rigorous, non-technical understanding of agentic and autonomous AI systems as they appear in real organizational workflows. It explains what makes a system agentic, how autonomy emerges from the interaction of models, planning, memory, tools, and feedback, and why these systems introduce risks far beyond predictive or assistive AI.
Participants examine how agentic systems act over time, interact with real environments and other systems, and fail through mechanisms such as runaway execution, silent objective drift, loss of auditability, and erosion of human accountability. The course is grounded in decision theory, planning, and control, but framed entirely at the system and governance level.
The emphasis is on autonomy boundaries, delegation risk, oversight failures, and organizational liability. Participants leave with the ability to critically assess vendor claims about “AI agents” and “autonomous workflows” and to design principled oversight and governance structures before deployment.
Core Topics:
What makes a system agentic versus predictive or assistive
From models to systems, planners, memory, tools, and execution
Planning and decision-making over time
Memory, state persistence, and compounding error
Tool use and real-world action
Feedback loops and self-reinforcement
Multi-agent and coordinated systems
Autonomy boundaries and delegation risk
Enterprise failure modes and accountability gaps
Illusions of control in agentic systems
Evaluation challenges for autonomous systems
Safety, risk, and governance for agentic AI
Where agentic systems add value and where they do not
Outcomes:
Explain what makes an AI system agentic rather than assistive
Understand how planning, memory, tools, and feedback create autonomy
Identify unintentional emergence of autonomy in workflows
Recognize failure modes unique to agentic and autonomous systems
Define appropriate autonomy boundaries for organizational use
Evaluate risks related to delegation and loss of human control
Critically assess vendor claims about AI agents and autonomy
Design oversight and governance principles for agentic systems
