FREE COURSE / SELF-PACED LEARNING

Detection Engineering for AI and Agentic Systems

Trace delegated authority, tool calls and evidence boundaries in AI agents with this free detection engineering course.

Read freely. Create an account to save your completed lessons.

Experience level
Advanced
Curriculum
3 modules · 3 lessons
Estimated reading
12 min
Language
English

Reading estimates use lesson text and an allowance for diagrams. Reflection, exercises and optional linked resources take additional time. These are estimates, not measured completion times.

KNOW WHAT YOU’RE WORKING TOWARD

What you’ll learn

  • Map an AI agent's delegated authority and data-access boundaries.
  • Select bounded tool-call evidence that can distinguish permitted work from an unexpected disclosure path.

Is this the right course for you?

Who it’s for

Security practitioners assessing AI assistants and agent workflows.

Scope and boundaries

Evidence and authority boundaries in agentic systems. Examples are conceptual and do not provide access to a production AI environment.

How you’ll learn

Self-paced reading with explanatory diagrams and linked resources. Reflection and optional resource study take additional time.

Before you begin

  • Detection Engineering: From Telemetry to Tuned Detections — You should understand the basic detection lifecycle and common telemetry sources.

About this course

Understand how identity, delegated authority, tool use, memory, and data movement create evidence in AI systems that can act on a user’s behalf. The course is written for one reader and develops its ideas through continuous explanation, realistic thought experiments, and explicit distinctions between fact, assessment, and uncertainty.

TRY THE REASONING / FICTIONAL SCENARIO

A tool call outside the user's request

An assistant was asked to summarize an internal report. Its run record contains a later call to send that report to an external destination. The tool was available, but the user did not request delivery.

Your task

Separate tool availability, user intent and actual delivery. Identify evidence needed to establish the effect without collecting unnecessary report contents.

Compare your reasoning

Tool availability is a capability, not proof of authorization for this request. Compare the request, delegated policy, tool-call arguments and effect receipt. A requested send is not equivalent to completed delivery. Preserve scoped identifiers and authorization context while avoiding unnecessary copies of sensitive content. Keep claims about cause and impact bounded by the records.

This example is a reflection exercise, not a scored assessment.

FOLLOW THE LEARNING PATH

Course curriculum

BUILD ON THE RIGHT FOUNDATION

These courses address different scopes. Choose the foundation or follow-on that fits your next question.

Sources and editorial review

Reviewed by Threat Intelligence Lab

Sources linked in the lessons provide context and further reading. Check the original documentation for current platform behavior and requirements.

Selected references are shown here; the lessons contain the full reading lists.

About this learning content: Courses, lessons, assessments, explanations and illustrations may be created with the help of artificial intelligence. We review and check the material and do our best to avoid incorrect or outdated information, but mistakes, omissions or ambiguous questions may remain. Please verify information before relying on it for professional, security, legal or operational decisions. Read the full notice or report an issue.