Pass IAPP AIGP Exam in First Attempt Guaranteed!
Get 100% Latest Exam Questions, Accurate & Verified Answers!
30 Days Free Updates, Instant Download!
IAPP AIGP Practice Test 47
Question 460
A disaster-response risk model outputs probabilities, yet events predicted at 90 percent occur only half the time. Which additional observation would most strongly confirm the particular risk or control gap?
Explanation
Correct Answer
Reliability diagrams show severe mismatch between forecast bands and observed frequency.
Incorrect Answers
Dataset profiling reveals that rural utterances represent a tiny share of the intended population. Reliability diagrams show severe mismatch between forecast bands and observed frequency.
The broker provides a file download but no evidence of collection rights or individual notice. Reliability diagrams show severe mismatch between forecast bands and observed frequency.
Identity matches show related records from one person in both train and test partitions. Reliability diagrams show severe mismatch between forecast bands and observed frequency.
Question 461
An energy provider's data science team approves a customer forecasting model without consulting safety, accessibility or customer-service teams. Select the ONE intervention-and-verification pair that directly resolves this specific governance gap.
Explanation
Correct Answer
Intervention: Create a cross-functional review with technical, legal, domain and affected-user expertise at design gates. / Verification: Approval packets include the required disciplines and recorded resolution of raised issues.
Incorrect Answers
Intervention: Publish a multi-function AI incident route with severity criteria, communications owners and response authority. / Verification: Approval packets include the required disciplines and recorded resolution of raised issues. Intervention: Create a cross-functional review with technical, legal, domain and affected-user expertise at design gates. / Verification: Approval packets include the required disciplines and recorded resolution of raised issues.
Intervention: Create a cross-functional review with technical, legal, domain and affected-user expertise at design gates. / Verification: A role matrix and contract assign each model-level and use-site control to a responsible party. Intervention: Create a cross-functional review with technical, legal, domain and affected-user expertise at design gates. / Verification: Approval packets include the required disciplines and recorded resolution of raised issues.
Intervention: Provide concise risk appetite, incident trend, exceptions and required decision reports tailored to directors. / Verification: Decision papers connect open risks to a responsible executive and an explicit accepted response. Intervention: Create a cross-functional review with technical, legal, domain and affected-user expertise at design gates. / Verification: Approval packets include the required disciplines and recorded resolution of raised issues.
Question 462
An insurer reports excellent risk prediction, but records from the same applicant appear in both training and evaluation datasets. Which governance reasoning best supports the appropriate response?
Explanation
Correct Answer
Overlapping training and test information can produce falsely optimistic performance claims. This reasoning fits train-test leakage by customer identity because the observed gap is that identity matches show related records from one person in both train and test partitions. It points to the practical intervention: Split and audit data at the appropriate individual level to prevent leakage. The outcome should then show that a leakage-free holdout evaluation reports realistic generalization to unseen applicants.
Incorrect Answers
Label quality shapes what supervised learning systems can reliably learn and be evaluated against. This reasoning fits train-test leakage by customer identity because the observed gap is that identity matches show related records from one person in both train and test partitions. Overlapping training and test information can produce falsely optimistic performance claims.
Data provenance supports reproducibility, investigation and accountability across AI pipelines. This reasoning fits train-test leakage by customer identity because the observed gap is that identity matches show related records from one person in both train and test partitions. Overlapping training and test information can produce falsely optimistic performance claims.
Probabilities used in governance decisions need calibration, not just an attractive ranking score. This reasoning fits train-test leakage by customer identity because the observed gap is that identity matches show related records from one person in both train and test partitions. Overlapping training and test information can produce falsely optimistic performance claims.
Question 463
A hospital agrees to an AI dictation service SLA based solely on API uptime, despite errors that alter clinical terms. Which governance reasoning best supports the appropriate response?
Explanation
Correct Answer
Uptime alone does not show that an AI system is safe or useful for a given task. This reasoning fits service-level promises and model degradation because the observed gap is that the service is reachable while domain-critical medication names are mis-transcribed. It points to the practical intervention: Negotiate task-appropriate quality measures, failure handling and support commitments in addition to availability. The outcome should then show that representative quality tests and escalation paths satisfy the agreed clinical workflow limits.
Incorrect Answers
Data residency must cover operational logs and subcontractors, not only primary inference. The outcome should then show that representative quality tests and escalation paths satisfy the agreed clinical workflow limits. Uptime alone does not show that an AI system is safe or useful for a given task.
AI vendor dependence creates continuity and switching risks that need advance treatment. The outcome should then show that representative quality tests and escalation paths satisfy the agreed clinical workflow limits. Uptime alone does not show that an AI system is safe or useful for a given task.
A supplier's assessment cannot fully substitute for a deployer's contextual impact review. The outcome should then show that representative quality tests and escalation paths satisfy the agreed clinical workflow limits. Uptime alone does not show that an AI system is safe or useful for a given task.
Question 464
A travel platform's model provider silently changes safety behavior while the customer's regression checks only test availability. Which test result would best demonstrate that the actual governance concern has been addressed?
Explanation
Correct Answer
An approved provider update passes behavioral tests and can be rolled back or contained. The required intervention is to require release notes, version pinning where supported and customer regression tests for provider changes. The particular diagnostic issue is that safety regressions coincide with a vendor version swap absent from internal change logs. The governing distinction is that third-party AI changes create a release risk even when the integrating app code is unchanged.
Incorrect Answers
The incident scenario no longer reproduces and regression tests document risk reduction. Third-party AI changes create a release risk even when the integrating app code is unchanged. An approved provider update passes behavioral tests and can be rolled back or contained.
The approved release record links version, tests, known limitations and responsible owners. Third-party AI changes create a release risk even when the integrating app code is unchanged. An approved provider update passes behavioral tests and can be rolled back or contained.
Threshold breaches produce timely alerts and tested containment decisions. Third-party AI changes create a release risk even when the integrating app code is unchanged. An approved provider update passes behavioral tests and can be rolled back or contained.
Question 465
A marketing model predicts likely pregnancy from shopping behavior without explicitly collecting a health attribute. Which additional observation would most strongly confirm the particular risk or control gap?
Explanation
Correct Answer
Derived scores strongly reveal sensitive health states despite no health field being provided.
Incorrect Answers
Test results use only high-resolution well-lit images unlike field conditions. Derived scores strongly reveal sensitive health states despite no health field being provided.
The team cannot reproduce how the feature used in a disputed decision was calculated. Derived scores strongly reveal sensitive health states despite no health field being provided.
Reliability diagrams show severe mismatch between forecast bands and observed frequency. Derived scores strongly reveal sensitive health states despite no health field being provided.
Question 466
A bank deploys a suspicious-transaction assistant; analysts accept every flag because they have not been trained on model limits. Which governance reasoning best supports the appropriate response?
Explanation
Correct Answer
User competence is an ongoing control for safe AI deployment. This reasoning fits user training on human override because the observed gap is that pilot audits show analysts treat a low-confidence output as a final mandatory decision. It points to the practical intervention: Train analysts on uncertainty, review duties, override steps and escalation criteria. The outcome should then show that training scenarios demonstrate correct overrides, rationale capture and escalation.
Incorrect Answers
Prohibition without usable alternatives can displace rather than eliminate AI risk. This reasoning fits user training on human override because the observed gap is that pilot audits show analysts treat a low-confidence output as a final mandatory decision. User competence is an ongoing control for safe AI deployment.
Operational success depends on the impact of predictions, not merely uptime. This reasoning fits user training on human override because the observed gap is that pilot audits show analysts treat a low-confidence output as a final mandatory decision. User competence is an ongoing control for safe AI deployment.
AI incident severity should reflect potential impact, not just software error type. This reasoning fits user training on human override because the observed gap is that pilot audits show analysts treat a low-confidence output as a final mandatory decision. User competence is an ongoing control for safe AI deployment.
Question 467
A design agency plans to sell generated illustrations without reviewing whether source asset licenses permit the intended commercialization. Which test result would best demonstrate that the actual governance concern has been addressed?
Explanation
Correct Answer
The approved content workflow documents legal use rights for all material inputs and planned distribution. Generative output does not automatically erase rights and restrictions attached to underlying assets.
Incorrect Answers
The insurer documents permitted rating factors and tests for prohibited proxy effects. Generative output does not automatically erase rights and restrictions attached to underlying assets. The approved content workflow documents legal use rights for all material inputs and planned distribution.
Credit decisions satisfy applicable fairness and adverse-action explanation controls after testing. Generative output does not automatically erase rights and restrictions attached to underlying assets. The approved content workflow documents legal use rights for all material inputs and planned distribution.
Safety testing and corrective controls address foreseeable failure modes before distribution. Generative output does not automatically erase rights and restrictions attached to underlying assets. The approved content workflow documents legal use rights for all material inputs and planned distribution.
Question 468
A public transport authority plans to release an arrival-time predictor with no documented intended users, limitations or evaluation. Which governance intervention most directly addresses this specific situation?
Explanation
Correct Answer
Prepare a release-ready model or system card detailing purpose, data, performance, harms and usage constraints. The action addresses release readiness model card rather than treating every AI control as interchangeable. The decisive signal to verify is that operations can access binary artifacts but have no approved explanation of tested contexts or limitations. This matters because model documentation supports transparent, accountable transition from development to use. Success should be measured by confirming that the approved release record links version, tests, known limitations and responsible owners.
Incorrect Answers
Use an evidence-driven maintenance schedule with material-change triggers and acceptance checks. Model documentation supports transparent, accountable transition from development to use. Prepare a release-ready model or system card detailing purpose, data, performance, harms and usage constraints.
Monitor outcome drift relative to changed policy and initiate governed revalidation. Model documentation supports transparent, accountable transition from development to use. Prepare a release-ready model or system card detailing purpose, data, performance, harms and usage constraints.
Apply a formal release gate with tolerances, remediation requirements and stop authority. Model documentation supports transparent, accountable transition from development to use. Prepare a release-ready model or system card detailing purpose, data, performance, harms and usage constraints.
Question 469
A media agency buys image-generation access and assumes every output is exclusively owned and free of third-party claims. Which governance reasoning best supports the appropriate response?
Explanation
Correct Answer
An API subscription does not automatically resolve generated-content ownership and liability. This reasoning fits licensing for generated content because the observed gap is that the terms grant use rights but explicitly exclude certain warranties and ownership assurances. It points to the practical intervention: Review output-use licenses, intellectual-property risks and indemnity scope before commercialization. The outcome should then show that the business has a rights and disclosure decision suited to its intended distribution.
Incorrect Answers
Contract risk allocation may depend on how the customer configures and uses the AI system. This reasoning fits licensing for generated content because the observed gap is that the terms grant use rights but explicitly exclude certain warranties and ownership assurances. An API subscription does not automatically resolve generated-content ownership and liability.
Vendor incident terms must cover security and AI-specific harm response needs. This reasoning fits licensing for generated content because the observed gap is that the terms grant use rights but explicitly exclude certain warranties and ownership assurances. An API subscription does not automatically resolve generated-content ownership and liability.
Building AI in-house can increase direct accountability rather than eliminate risk. This reasoning fits licensing for generated content because the observed gap is that the terms grant use rights but explicitly exclude certain warranties and ownership assurances. An API subscription does not automatically resolve generated-content ownership and liability.
Question 470
A health insurer's AI segments customers into price groups but management cannot demonstrate the absence of unlawful rating factors. Which governance intervention most directly addresses this specific situation?
Explanation
Correct Answer
Review insurance fairness rules and model factors with legal and actuarial oversight. The action addresses insurance pricing transparency rather than treating every AI control as interchangeable. The decisive signal to verify is that pricing models rely on opaque correlations that cannot be linked to permitted risk variables. This matters because an insurance model's technical accuracy does not guarantee its rates are legally permissible. Success should be measured by confirming that the insurer documents permitted rating factors and tests for prohibited proxy effects.
Incorrect Answers
Review lending nondiscrimination and decision-notice obligations appropriate to the jurisdiction. An insurance model's technical accuracy does not guarantee its rates are legally permissible. Review insurance fairness rules and model factors with legal and actuarial oversight.
Review confidentiality, trade-secret protection and vendor use restrictions before disclosure. An insurance model's technical accuracy does not guarantee its rates are legally permissible. Review insurance fairness rules and model factors with legal and actuarial oversight.
Assess copyright and license permissions and remove or remediate unauthorized training use. An insurance model's technical accuracy does not guarantee its rates are legally permissible. Review insurance fairness rules and model factors with legal and actuarial oversight.
Artificial Intelligence Governance Professional (AIGP) Practice Tests
- IAPP AIGP Practice Test 45 - 10 Questions (Last Reviewed 08-Oct-2026)
- IAPP AIGP Practice Test 46 - 10 Questions (Last Reviewed 07-Oct-2026)
- IAPP AIGP Practice Test 47 - 10 Questions (Last Reviewed 06-Oct-2026)
- IAPP AIGP Practice Test 48 - 10 Questions (Last Reviewed 05-Oct-2026)
- IAPP AIGP Practice Test 49 - 10 Questions (Last Reviewed 04-Oct-2026)
