Study GuidesAIGP Study Guide

Apply what you study

AIGP scenario workshop

Explain why an answer fits the facts. Choose a response, identify the decisive fact, then compare every choice.

Four original teaching examples. Free, with no account needed. These are independent practice scenarios, not IAPP exam questions.

Sources checked October 20, 2026. These cases teach selected distinctions and do not cover the complete exam outline.

A method for unfamiliar questions

  1. Identify the use case, affected people and life cycle stage.
  2. Identify the actor and applicable law or framework.
  3. Choose a control that addresses the stated risk.
  4. Name the owner, evidence and monitoring needed to make it work.

NIST functions

GOVERN establishes governance. MAP establishes context. MEASURE assesses risk. MANAGE prioritizes and responds.

Frameworks and law

The NIST AI RMF is voluntary. It does not replace applicable legal duties.

Average accuracy

An overall score can conceal poor results for particular groups or operating conditions.

Deployment approval

An approval is a decision with owners and conditions. Monitoring continues after deployment.

01 · NIST AI RMF

Understand the setting first

A team documents an AI system's intended use, affected people, deployment setting and potential impacts. Which NIST AI RMF function best fits?

Choose the strongest answer

Read the worked explanation

The decisive fact

The team is establishing context and identifying impacts. That is the central work of MAP.

Why each choice works or fails

  1. Correct. MAP frames the system, context and risks so later measurement and management are meaningful.
  2. MEASURE uses analysis, assessment and testing to evaluate risks. The facts describe the earlier context work.
  3. MANAGE prioritizes and responds to risks. The scenario has not yet described that response.

What would change the answer

If the team tested performance and evaluated identified risks, MEASURE would be the better fit.

NIST AI RMF 1.0, Part 2, MAP function. Voluntary framework. NIST AI RMF 1.0, functions and categories.

Study the related lesson

02 · Accountability

Nobody owns the decision

Teams disagree about who may approve AI deployment and who owns incident escalation. Which response best addresses the gap?

Choose the strongest answer

Read the worked explanation

The decisive fact

The problem is missing responsibility and authority. A technical metric cannot assign an accountable owner.

Why each choice works or fails

  1. Accuracy may matter, but it does not resolve approval authority or escalation responsibility.
  2. Correct. GOVERN establishes policies, roles and accountability across the other functions and the AI life cycle.
  3. A published score does not decide who accepts risk or responds to an incident.

What would change the answer

If responsibilities were clear but their controls failed, the next step would focus on testing and improving those controls.

NIST AI RMF 1.0, GOVERN 2. Roles, responsibilities and communication. NIST AI RMF 1.0, functions and categories.

Study the related lesson

03 · Evaluation

A high average hides uneven results

An AI evaluation reports 95 percent overall accuracy, but one affected group has a much higher error rate. What should the governance review do?

Choose the strongest answer

Read the worked explanation

The decisive fact

An aggregate result does not establish acceptable performance for every affected group. Review the relevant context and harms.

Why each choice works or fails

  1. A high average does not resolve uneven performance or its consequences.
  2. A larger aggregate does not make a materially affected group irrelevant.
  3. Correct. Examine test validity, relevant group performance and impact. Choose controls and document any remaining risk.

What would change the answer

Even if group metrics were similar, testing would still need to address validity, reliability, safety and the actual use context.

NIST AI RMF 1.0, MEASURE 2.11 and MAP 5. Bias, fairness and impacts. NIST AI RMF 1.0, functions and categories.

Study the related lesson

04 · Deployment monitoring

The production setting changes

After deployment, users and incoming data change substantially. The original evaluation no longer represents current use. What is the best next action?

Choose the strongest answer

Read the worked explanation

The decisive fact

The deployment context has changed. Governance needs evidence from actual operation, along with a response to new or changed risk.

Why each choice works or fails

  1. Correct. Monitor and reassess the system. Document ownership, response thresholds and whether continued use remains appropriate.
  2. An earlier approval does not establish continued suitability under changed conditions.
  3. User choice does not replace evaluation, risk management or applicable legal obligations.

What would change the answer

A stable setting still needs proportionate ongoing monitoring. A material change makes reassessment especially important.

NIST AI RMF 1.0, MEASURE 2.4 and MANAGE 4. Monitoring and risk response. NIST AI RMF 1.0, functions and categories.

Study the related lesson

Turn a mistake into a study task

  1. Name the fact you missed.
  2. Write the rule in one sentence.
  3. Explain why your chosen answer fails.
  4. Read the related lesson, then answer again without looking.
Plan your next study session

Keep the useful part of a mistake

Know what to review next.

Record the rule you missed, the decisive fact and your next review date. Free, with printable and downloadable templates.

Open the mistake review log