IAPP AIGP Practice Test 37 - Prepaway
Practice Exams:

Pass IAPP AIGP Exam in First Attempt Guaranteed!

Get 100% Latest Exam Questions, Accurate & Verified Answers!

30 Days Free Updates, Instant Download!

Verified By Experts
Premium Bundle
$29.98
$79.99 $109.97

AIGP Premium Bundle


  • Premium File 222 Questions & Answers. Last update: Nov 15, 2026
  • Training Course 124 Lectures
  • Study Guide 566 Pages

IAPP AIGP Practice Test 37

Question 360

A retailer routes EU customer-support prompts to a service outside the approved region with no transfer review. Which additional observation would most strongly confirm the particular risk or control gap?

  1. Network logs show personal prompts leave the expected territory through an undocumented inference endpoint.
  2. The vendor notice exists, but the lender has no record explaining the legal basis for its distinct profiling activity.
  3. A transcript shows identifiable data disclosed to an unauthorized user and no breach triage record.
  4. Reject decisions would become final without a person having authority to reconsider a case.

Explanation

Correct Answer

A:

Network logs show personal prompts leave the expected territory through an undocumented inference endpoint.

Incorrect Answers

B:

The vendor notice exists, but the lender has no record explaining the legal basis for its distinct profiling activity. Network logs show personal prompts leave the expected territory through an undocumented inference endpoint.

C:

A transcript shows identifiable data disclosed to an unauthorized user and no breach triage record. Network logs show personal prompts leave the expected territory through an undocumented inference endpoint.

D:

Reject decisions would become final without a person having authority to reconsider a case. Network logs show personal prompts leave the expected territory through an undocumented inference endpoint.

Question 361

A consumer app vendor claims its new AI moderation is safe based on a hand-picked demonstration of 20 prompts. Which governance reasoning best supports the appropriate response?

  1. Model confidence and scope limits must influence system decisions, not merely appear in an appendix.
  2. AI threat models must include attacks on data and model behavior, not only network intrusion.
  3. Evidence from curated demos does not establish robustness across deployment conditions.
  4. AI impact assessment links a business use case to consequences for affected people.

Explanation

Correct Answer

C:

Evidence from curated demos does not establish robustness across deployment conditions. This reasoning fits benchmark contamination and pilots because the observed gap is that the shown prompts avoid ambiguous, multilingual and adversarial inputs that real customers send. It points to the practical intervention: Design representative benchmarks and controlled pilot tests before approving rollout. The outcome should then show that the pilot records performance under realistic difficult cases and meets pre-agreed guardrail criteria.

Incorrect Answers

A:

Model confidence and scope limits must influence system decisions, not merely appear in an appendix. This reasoning fits benchmark contamination and pilots because the observed gap is that the shown prompts avoid ambiguous, multilingual and adversarial inputs that real customers send. Evidence from curated demos does not establish robustness across deployment conditions.

B:

AI threat models must include attacks on data and model behavior, not only network intrusion. This reasoning fits benchmark contamination and pilots because the observed gap is that the shown prompts avoid ambiguous, multilingual and adversarial inputs that real customers send. Evidence from curated demos does not establish robustness across deployment conditions.

D:

AI impact assessment links a business use case to consequences for affected people. This reasoning fits benchmark contamination and pilots because the observed gap is that the shown prompts avoid ambiguous, multilingual and adversarial inputs that real customers send. Evidence from curated demos does not establish robustness across deployment conditions.

Question 362

A transport app adds a voice-only AI journey planner, excluding passengers who cannot hear spoken responses. Which additional observation would most strongly confirm the particular risk or control gap?

  1. The prototype has no equivalent text controls or captions for essential instructions.
  2. Usability criteria reflect sales expectations but no child safeguarding or classroom context.
  3. The proposed pipeline treats anonymous labels as authoritative without provenance checks.
  4. The feature store includes sensitive attributes with no decision or documented necessity.

Explanation

Correct Answer

A:

The prototype has no equivalent text controls or captions for essential instructions. A meaningful follow-up would be to verify that accessibility testing shows equivalent core journey tasks can be completed without hearing speech.

Incorrect Answers

B:

Usability criteria reflect sales expectations but no child safeguarding or classroom context. A meaningful follow-up would be to verify that accessibility testing shows equivalent core journey tasks can be completed without hearing speech. The prototype has no equivalent text controls or captions for essential instructions.

C:

The proposed pipeline treats anonymous labels as authoritative without provenance checks. A meaningful follow-up would be to verify that accessibility testing shows equivalent core journey tasks can be completed without hearing speech. The prototype has no equivalent text controls or captions for essential instructions.

D:

The feature store includes sensitive attributes with no decision or documented necessity. A meaningful follow-up would be to verify that accessibility testing shows equivalent core journey tasks can be completed without hearing speech. The prototype has no equivalent text controls or captions for essential instructions.

Question 363

A city office wants AI to assign parking-permit renewals even though a deterministic rule engine already handles all documented exceptions. Which test result would best demonstrate that the actual governance concern has been addressed?

  1. Threat and access tests demonstrate employee information is restricted to authorized purposes.
  2. A documented alternatives assessment selects an appropriately limited solution with equal or better service outcomes.
  3. A documented override route prevents contested cancellations without appropriate authorization.
  4. The design review documents input and mitigation decisions for each materially affected group.

Explanation

Correct Answer

B:

A documented alternatives assessment selects an appropriately limited solution with equal or better service outcomes. AI deployment should be justified by the problem and expected benefit, not by the availability of a model.

Incorrect Answers

A:

Threat and access tests demonstrate employee information is restricted to authorized purposes. AI deployment should be justified by the problem and expected benefit, not by the availability of a model. A documented alternatives assessment selects an appropriately limited solution with equal or better service outcomes.

C:

A documented override route prevents contested cancellations without appropriate authorization. AI deployment should be justified by the problem and expected benefit, not by the availability of a model. A documented alternatives assessment selects an appropriately limited solution with equal or better service outcomes.

D:

The design review documents input and mitigation decisions for each materially affected group. AI deployment should be justified by the problem and expected benefit, not by the availability of a model. A documented alternatives assessment selects an appropriately limited solution with equal or better service outcomes.

Question 364

A pharmacy uses an AI prescription-routing tool and has no procedure when inference fails mid-shift. Which test result would best demonstrate that the actual governance concern has been addressed?

  1. Authorized staff can continue critical processing without using unverified model outputs.
  2. Security tests show the assistant cannot invoke unauthorized actions or access unrelated records.
  3. Security exercises confirm permissions and approval gates resist manipulation.
  4. A test appeal reaches an authorized reviewer and resolves documented factual errors.

Explanation

Correct Answer

A:

Authorized staff can continue critical processing without using unverified model outputs. The required intervention is to create a safe human fallback with adequate staffing, queues and incident escalation. The particular diagnostic issue is that when the model endpoint fails, prescriptions remain in an unattended electronic queue. The governing distinction is that aI operational resilience includes non-AI continuity for essential services.

Incorrect Answers

B:

Security tests show the assistant cannot invoke unauthorized actions or access unrelated records. AI operational resilience includes non-AI continuity for essential services. Authorized staff can continue critical processing without using unverified model outputs.

C:

Security exercises confirm permissions and approval gates resist manipulation. AI operational resilience includes non-AI continuity for essential services. Authorized staff can continue critical processing without using unverified model outputs.

D:

A test appeal reaches an authorized reviewer and resolves documented factual errors. AI operational resilience includes non-AI continuity for essential services. Authorized staff can continue critical processing without using unverified model outputs.

Question 365

An architecture studio has no rule on whether employees may ask an AI tool to generate confidential client plans. Which governance intervention most directly addresses this specific situation?

  1. Add provenance, lawful-use warranties, incident reporting and audit cooperation to the vendor review.
  2. Add approved-subprocessor disclosure, flow-down controls and change notification to procurement processes.
  3. Create an AI acceptable-use policy describing allowed data, tool approvals and professional review.
  4. Add fairness, accessibility and contestability acceptance criteria to design requirements.

Explanation

Correct Answer

C:

Create an AI acceptable-use policy describing allowed data, tool approvals and professional review. Rules for employee AI usage must update existing confidentiality and data security controls.

Incorrect Answers

A:

Add provenance, lawful-use warranties, incident reporting and audit cooperation to the vendor review. Rules for employee AI usage must update existing confidentiality and data security controls. Create an AI acceptable-use policy describing allowed data, tool approvals and professional review.

B:

Add approved-subprocessor disclosure, flow-down controls and change notification to procurement processes. Rules for employee AI usage must update existing confidentiality and data security controls. Create an AI acceptable-use policy describing allowed data, tool approvals and professional review.

D:

Add fairness, accessibility and contestability acceptance criteria to design requirements. Rules for employee AI usage must update existing confidentiality and data security controls. Create an AI acceptable-use policy describing allowed data, tool approvals and professional review.

Question 366

A global shipping firm distributes a one-page AI policy but frontline dispatchers keep uploading live manifests to unsanctioned chatbots. Which additional observation would most strongly confirm the particular risk or control gap?

  1. Escalations bounce between technology, legal and product without a person accountable for residual risk.
  2. Incident reports show staff misunderstand which shipment fields may enter third-party AI services.
  3. Local teams cannot document country-specific impact requirements in the headquarters-only template.
  4. The customer handbook omits warnings that the provider lists about unsupported document types.

Explanation

Correct Answer

B:

Incident reports show staff misunderstand which shipment fields may enter third-party AI services.

Incorrect Answers

A:

Escalations bounce between technology, legal and product without a person accountable for residual risk. Incident reports show staff misunderstand which shipment fields may enter third-party AI services.

C:

Local teams cannot document country-specific impact requirements in the headquarters-only template. Incident reports show staff misunderstand which shipment fields may enter third-party AI services.

D:

The customer handbook omits warnings that the provider lists about unsupported document types. Incident reports show staff misunderstand which shipment fields may enter third-party AI services.

Question 367

A disaster-response risk model outputs probabilities, yet events predicted at 90 percent occur only half the time. Which governance intervention most directly addresses this specific situation?

  1. Measure calibration alongside discrimination and correct or qualify probability estimates.
  2. Stratify and validate training and test samples for the actual served language communities.
  3. Establish lawful rights and permitted use for each data source before training.
  4. Split and audit data at the appropriate individual level to prevent leakage.

Explanation

Correct Answer

A:

Measure calibration alongside discrimination and correct or qualify probability estimates. The action addresses evaluation of calibration and uncertainty rather than treating every AI control as interchangeable. The decisive signal to verify is that reliability diagrams show severe mismatch between forecast bands and observed frequency. This matters because probabilities used in governance decisions need calibration, not just an attractive ranking score. Success should be measured by confirming that post-validation probabilities are calibrated for the intended decision context or clearly limited.

Incorrect Answers

B:

Stratify and validate training and test samples for the actual served language communities. Probabilities used in governance decisions need calibration, not just an attractive ranking score. Measure calibration alongside discrimination and correct or qualify probability estimates.

C:

Establish lawful rights and permitted use for each data source before training. Probabilities used in governance decisions need calibration, not just an attractive ranking score. Measure calibration alongside discrimination and correct or qualify probability estimates.

D:

Split and audit data at the appropriate individual level to prevent leakage. Probabilities used in governance decisions need calibration, not just an attractive ranking score. Measure calibration alongside discrimination and correct or qualify probability estimates.

Question 368

A utility's emissions prediction model shows stable averages, but no independent review tests edge cases or data integrity. Which governance intervention most directly addresses this specific situation?

  1. Maintain validated rollback artifacts and a rehearsed controlled reversion procedure.
  2. Classify and document AI harms using an incident taxonomy with escalation and reporting review.
  3. Reassess risk, owners and residual exposure whenever material features or sources change.
  4. Schedule evidence-based audits of model quality, validity and safety assumptions.

Explanation

Correct Answer

D:

Schedule evidence-based audits of model quality, validity and safety assumptions. The action addresses periodic model quality audit rather than treating every AI control as interchangeable. The decisive signal to verify is that operations reports one aggregate mean while unusual peak-emission periods are not reviewed. This matters because periodic assurance must challenge the conditions and metrics used to declare a model reliable. Success should be measured by confirming that audits cover critical subpopulations and exceptional conditions with traceable findings.

Incorrect Answers

A:

Maintain validated rollback artifacts and a rehearsed controlled reversion procedure. Periodic assurance must challenge the conditions and metrics used to declare a model reliable. Schedule evidence-based audits of model quality, validity and safety assumptions.

B:

Classify and document AI harms using an incident taxonomy with escalation and reporting review. Periodic assurance must challenge the conditions and metrics used to declare a model reliable. Schedule evidence-based audits of model quality, validity and safety assumptions.

C:

Reassess risk, owners and residual exposure whenever material features or sources change. Periodic assurance must challenge the conditions and metrics used to declare a model reliable. Schedule evidence-based audits of model quality, validity and safety assumptions.

Question 369

A logistics operator plans to use image recognition to detect unsafe forklift zones without retaining manual alarms. Which test result would best demonstrate that the actual governance concern has been addressed?

  1. Subgroup checks and documented remediation meet the approved nondiscrimination thresholds.
  2. The process remains safe when the vision model fails, with measured false-negative rates under threshold.
  3. Students can request review, correct relevant facts and receive a recorded decision from an authorized reviewer.
  4. The released assistant meets approved service-quality floors for each supported dialect group.

Explanation

Correct Answer

B:

The process remains safe when the vision model fails, with measured false-negative rates under threshold. AI performance uncertainty requires safety design beyond ordinary average accuracy.

Incorrect Answers

A:

Subgroup checks and documented remediation meet the approved nondiscrimination thresholds. AI performance uncertainty requires safety design beyond ordinary average accuracy. The process remains safe when the vision model fails, with measured false-negative rates under threshold.

C:

Students can request review, correct relevant facts and receive a recorded decision from an authorized reviewer. AI performance uncertainty requires safety design beyond ordinary average accuracy. The process remains safe when the vision model fails, with measured false-negative rates under threshold.

D:

The released assistant meets approved service-quality floors for each supported dialect group. AI performance uncertainty requires safety design beyond ordinary average accuracy. The process remains safe when the vision model fails, with measured false-negative rates under threshold.

Question 370

A financial fraud model drifts, but the infrastructure team keeps restarting servers despite a silent label-definition change. Select the ONE intervention-and-verification pair that directly resolves this specific governance gap.

  1. Intervention: Investigate drift with data, product and operational experts before changing infrastructure. / Verification: A corrected label mapping and evaluation restore model reliability without unnecessary restarts.
  2. Intervention: Investigate drift with data, product and operational experts before changing infrastructure. / Verification: A monitoring dashboard and remediation protocol detect and address material group differences.
  3. Intervention: Create accurate technical and operational disclosures linked to the released model version. / Verification: Downstream users receive current limitation and oversight guidance with the actual release.
  4. Intervention: Threat-model the artifact, build pipeline and inference dependencies and add integrity controls. / Verification: A corrected label mapping and evaluation restore model reliability without unnecessary restarts.

Explanation

Correct Answer

A:

Intervention: Investigate drift with data, product and operational experts before changing infrastructure. / Verification: A corrected label mapping and evaluation restore model reliability without unnecessary restarts.

Incorrect Answers

B:

Intervention: Investigate drift with data, product and operational experts before changing infrastructure. / Verification: A monitoring dashboard and remediation protocol detect and address material group differences. Intervention: Investigate drift with data, product and operational experts before changing infrastructure. / Verification: A corrected label mapping and evaluation restore model reliability without unnecessary restarts.

C:

Intervention: Create accurate technical and operational disclosures linked to the released model version. / Verification: Downstream users receive current limitation and oversight guidance with the actual release. Intervention: Investigate drift with data, product and operational experts before changing infrastructure. / Verification: A corrected label mapping and evaluation restore model reliability without unnecessary restarts.

D:

Intervention: Threat-model the artifact, build pipeline and inference dependencies and add integrity controls. / Verification: A corrected label mapping and evaluation restore model reliability without unnecessary restarts. Intervention: Investigate drift with data, product and operational experts before changing infrastructure. / Verification: A corrected label mapping and evaluation restore model reliability without unnecessary restarts.

Artificial Intelligence Governance Professional (AIGP) Practice Tests