FirstPass AIGP
Playbook Progress (0% Completed)

You are here: Practice › Mindset simulator

Balanced Mindset Simulator

Five workplace scenarios. Pick A, B or C, then read why the balanced answer works. Your score updates as you go.

Spot the balanced governance answer. (A) bans everything, (B) ignores risk, and (C) identifies, manages and mitigates risk in proportion.

Score: 0 of 0 balanced

Question traps: community notes (unverified)

A study note shared online describes the patterns below. We have not checked every point against the source texts. Treat each one as a prompt to look something up, not as a rule.

  • Role mix-up (provider vs deployer). The note suggests questions may test whether a company buying a ready-made tool is treated as the builder. It also suggests a deployer that changes a system's intended purpose may take on provider duties. Check the definitions in the EU AI Act and Article 25.
  • Automatic high-risk assumption. The note warns against labelling any system in a sensitive sector as high-risk without checking the exceptions. It lists four exceptions in Article 6(3), with profiling as a key limit. Check the article text, including the conditions.
  • Public data as a legal basis. The note says publicly available data is not, on its own, a legal basis under GDPR Article 6. It also points to Article 9(2)(e) for special category data that a person has manifestly made public. Read both articles and take legal advice for any real case.
  • Lifecycle and function mix-up. The note links the NIST AI RMF functions (Govern, Map, Measure, Manage) to typical activities, and suggests a first impact assessment usually comes early, at planning and design. Check the framework text and your own organisation's policy.
  • Near-synonyms. Terms such as interpretability and explainability, data drift and concept drift, provenance and lineage, and audit and assessment are often used loosely. Definitions vary between sources, so note the definition each question uses.
  • True but off-target answers. A correct statement can still be wrong for the question. Re-read the last line of the scenario to confirm the role, jurisdiction or lifecycle stage it asks about.
  • A popular tip, not an official rule. One widely shared tip says to pick the option that reduces risk and strengthens accountability. It can mislead when a question asks for a specific required step.

Practice questions we wrote

These are original questions written for this site in a similar style. Our readings are starting points. Confirm each one against the BoK v2.1 and the EU AI Act text before you rely on it.

1. A company buys a ready-made résumé-screening tool and uses it for its stated purpose. Who is most likely responsible for the technical documentation?
  • A. The company that deploys the tool
  • B. The vendor that built the tool
  • C. The applicants being screened
  • D. The national regulator

Our reading: B. The builder, which the Act calls the provider, is usually the party expected to prepare technical documentation. Check the provider duties in Article 16 and the documentation requirements in Annex IV.

2. When should a first impact assessment normally begin?
  • A. After launch, once real data exists
  • B. During planning and design, before the system is built
  • C. Only after a complaint is received
  • D. Only when the system is retired

Our reading: B. Early assessment can shape design choices. Your organisation's policy and the framework you follow may set the exact timing.

3. A fraud model becomes less accurate. Customer inputs look the same as before, but fraudsters have changed their tactics, so the link between inputs and fraud has shifted. Which term best describes this?
  • A. Data drift
  • B. Concept drift
  • C. A provenance gap
  • D. An interpretability gap

Our reading: B. The relationship between inputs and outcomes has changed. Data drift would mean the inputs themselves have shifted. Check the definitions your course uses.

4. A non-EU company supplies an AI system whose outputs are used by an EU organisation. Which principle best explains why EU rules might still apply?
  • A. Extraterritoriality
  • B. Reciprocity
  • C. Universal jurisdiction
  • D. Proportionality

Our reading: A. The Act's scope provisions can reach some providers and deployers outside the EU whose outputs are used in the EU. Check the scope article for the exact conditions.

5. A hiring tool only compiles applicant records for a recruiter and does not rank or reject anyone. Which statement is most accurate?
  • A. It is automatically high-risk because it is used in employment.
  • B. It may fall outside high-risk under the Article 6(3) exceptions, provided its conditions are met and it does not perform profiling.
  • C. It is not regulated under the Act in any form.
  • D. It is prohibited under the Act.

Our reading: B. The narrow procedural exception may apply, but the conditions must be checked, and profiling removes the exception. Read Article 6(3) in full.

Independent study guide. Not affiliated with, endorsed by or sponsored by the IAPP, Dr. Kyle David or Dr. David Privacy. Not legal advice. About this site
Copied to clipboard!