IAPP AIGP Practice Test 3
Question 20
A medical triage classifier returns confident-looking labels although its probability calibration differs sharply across patient groups. A separate assurance review finds the concern is relevant to an expanded user group. What is the most relevant fallback when the promised safeguard cannot be demonstrated?
- Route unreliable classifications to qualified human assessment while recalibrating and validating the model.
- Disable high-impact tool actions and revoke credentials if the agent exceeds its approved autonomy.
- Suspend unsupported automated adverse recommendations and route cases for explainable review.
- Stop the unapproved collection and reconsider privacy-preserving evaluation methods if proportionality cannot be shown.
Explanation
Correct Answer
A:Route unreliable classifications to qualified human assessment while recalibrating and validating the model. Correct. It limits further reliance or exposure while the risk remains unresolved. A probabilistic output is not a deterministic guarantee, especially in high-impact applications.
Incorrect Answers
B:Disable high-impact tool actions and revoke credentials if the agent exceeds its approved autonomy. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with Autonomy and agentic scope: Agentic AI requires controls over actions and delegated authority as well as generated text. The present case instead concerns Probabilistic uncertainty.
C:Suspend unsupported automated adverse recommendations and route cases for explainable review. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with AI opacity and explanation: Transparency and useful explainability must match the affected audience and impact, not simply display a score. The present case instead concerns Probabilistic uncertainty.
D:Stop the unapproved collection and reconsider privacy-preserving evaluation methods if proportionality cannot be shown. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with Responsible AI tradeoffs: Responsible AI principles interact; governance should record and justify risk tradeoffs rather than invoke a single slogan. The present case instead concerns Probabilistic uncertainty.
Question 21
A medical triage classifier returns confident-looking labels although its probability calibration differs sharply across patient groups. A material change to the operating model has made the earlier risk acceptance incomplete. What review step identifies the specific control gap rather than an unrelated AI risk?
- Map the agent's reachable tools, action permissions, delegation chain and irreversible side effects.
- Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning.
- Identify whether explanations describe case-specific factors, system limits and meaningful review routes.
- Inspect calibration curves and subgroup error rates, not just the displayed confidence percentage.
Explanation
Correct Answer
D:Inspect calibration curves and subgroup error rates, not just the displayed confidence percentage. Correct. It establishes the relevant factual risk boundary before implementation. A probabilistic output is not a deterministic guarantee, especially in high-impact applications.
Incorrect Answers
A:Map the agent's reachable tools, action permissions, delegation chain and irreversible side effects. Not selected. That check primarily diagnoses a different governance issue. It belongs with Autonomy and agentic scope: Agentic AI requires controls over actions and delegated authority as well as generated text. The present case instead concerns Probabilistic uncertainty.
B:Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning. Not selected. That check primarily diagnoses a different governance issue. It belongs with AI system boundary: AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. The present case instead concerns Probabilistic uncertainty.
C:Identify whether explanations describe case-specific factors, system limits and meaningful review routes. Not selected. That check primarily diagnoses a different governance issue. It belongs with AI opacity and explanation: Transparency and useful explainability must match the affected audience and impact, not simply display a score. The present case instead concerns Probabilistic uncertainty.
Question 22
A medical triage classifier returns confident-looking labels although its probability calibration differs sharply across patient groups. A material change to the operating model has made the earlier risk acceptance incomplete. Which safeguard most directly treats the described failure mode?
- Set confidence thresholds and an escalation pathway for uncertainty, with clinician review for consequential cases.
- Document the model, its inputs, predictions, decision points and accountable humans in the system inventory.
- Adopt suitable local explanations, decision records and a human challenge process without asserting that every feature is causal.
- Select a documented proportionate mitigation that tests fairness while minimizing sensitive-data access and securing any justified processing.
Explanation
Correct Answer
A:Set confidence thresholds and an escalation pathway for uncertainty, with clinician review for consequential cases. Correct. It directly reduces or governs the specific documented risk. A probabilistic output is not a deterministic guarantee, especially in high-impact applications.
Incorrect Answers
B:Document the model, its inputs, predictions, decision points and accountable humans in the system inventory. Not selected. That action treats another control objective, not the specific gap described. It belongs with AI system boundary: AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. The present case instead concerns Probabilistic uncertainty.
C:Adopt suitable local explanations, decision records and a human challenge process without asserting that every feature is causal. Not selected. That action treats another control objective, not the specific gap described. It belongs with AI opacity and explanation: Transparency and useful explainability must match the affected audience and impact, not simply display a score. The present case instead concerns Probabilistic uncertainty.
D:Select a documented proportionate mitigation that tests fairness while minimizing sensitive-data access and securing any justified processing. Not selected. That action treats another control objective, not the specific gap described. It belongs with Responsible AI tradeoffs: Responsible AI principles interact; governance should record and justify risk tradeoffs rather than invoke a single slogan. The present case instead concerns Probabilistic uncertainty.
Question 23
A medical triage classifier returns confident-looking labels although its probability calibration differs sharply across patient groups. A material change to the operating model has made the earlier risk acceptance incomplete. Which verification task should support the next approval decision?
- Attempt unauthorized and irreversible test actions and confirm the agent stops or requests approval.
- Test whether affected users and reviewers can understand and contest representative decisions.
- Measure calibration and false-negative performance by subgroup on representative evaluation data.
- Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning.
Explanation
Correct Answer
C:Measure calibration and false-negative performance by subgroup on representative evaluation data. Correct. It yields measurable evidence that the case-specific safeguard works. A probabilistic output is not a deterministic guarantee, especially in high-impact applications.
Incorrect Answers
A:Attempt unauthorized and irreversible test actions and confirm the agent stops or requests approval. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Autonomy and agentic scope: Agentic AI requires controls over actions and delegated authority as well as generated text. The present case instead concerns Probabilistic uncertainty.
B:Test whether affected users and reviewers can understand and contest representative decisions. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with AI opacity and explanation: Transparency and useful explainability must match the affected audience and impact, not simply display a score. The present case instead concerns Probabilistic uncertainty.
D:Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning. Not selected. This action relates to AI system boundary. AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. For the validate decision, it does not resolve the specifically described issue of Probabilistic uncertainty in A medical triage classifier returns confident-looking labels although its probability calibration differs sharply across patient groups. A material change to the operating model has made the earlier risk acceptance incomplete.
Question 24
A medical triage classifier returns confident-looking labels although its probability calibration differs sharply across patient groups. A material change to the operating model has made the earlier risk acceptance incomplete. Select TWO complementary actions: first assess the stated concern and then implement its targeted control.
- Set confidence thresholds and an escalation pathway for uncertainty, with clinician review for consequential cases.
- Inspect calibration curves and subgroup error rates, not just the displayed confidence percentage.
- Map the agent's reachable tools, action permissions, delegation chain and irreversible side effects.
- Document the model, its inputs, predictions, decision points and accountable humans in the system inventory.
Explanation
Correct Answers
A:Set confidence thresholds and an escalation pathway for uncertainty, with clinician review for consequential cases. Correct. As the second, implementation half of the requested pair, this directly applies the needed safeguard. A probabilistic output is not a deterministic guarantee, especially in high-impact applications.
B:Inspect calibration curves and subgroup error rates, not just the displayed confidence percentage. Correct. As the first, evidence-gathering half of the requested pair, this establishes the specific risk boundary. A probabilistic output is not a deterministic guarantee, especially in high-impact applications.
Incorrect Answers
C:Map the agent's reachable tools, action permissions, delegation chain and irreversible side effects. Not selected. It addresses a different governance mechanism, not the required assessment-and-control pair. It belongs with Autonomy and agentic scope: Agentic AI requires controls over actions and delegated authority as well as generated text. The present case instead concerns Probabilistic uncertainty.
D:Document the model, its inputs, predictions, decision points and accountable humans in the system inventory. Not selected. It addresses a different governance mechanism, not the required assessment-and-control pair. It belongs with AI system boundary: AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. The present case instead concerns Probabilistic uncertainty.
Question 25
A lender's ensemble recommendation is highly accurate overall but borrowers cannot understand why their cases were flagged. The system owner is requesting a documented approval of this existing governance concern. What review step identifies the specific control gap rather than an unrelated AI risk?
- Identify whether explanations describe case-specific factors, system limits and meaningful review routes.
- Estimate aggregate impact using exposure volume, severity and unequal subgroup effects.
- Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning.
- Map the agent's reachable tools, action permissions, delegation chain and irreversible side effects.
Explanation
Correct Answer
A:Identify whether explanations describe case-specific factors, system limits and meaningful review routes. Correct. It establishes the relevant factual risk boundary before implementation. Transparency and useful explainability must match the affected audience and impact, not simply display a score.
Incorrect Answers
B:Estimate aggregate impact using exposure volume, severity and unequal subgroup effects. Not selected. That check primarily diagnoses a different governance issue. It belongs with Scale-driven harm: Even a low individual error rate may create material harm when AI is applied at scale. The present case instead concerns AI opacity and explanation.
C:Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning. Not selected. That check primarily diagnoses a different governance issue. It belongs with AI system boundary: AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. The present case instead concerns AI opacity and explanation.
D:Map the agent's reachable tools, action permissions, delegation chain and irreversible side effects. Not selected. That check primarily diagnoses a different governance issue. It belongs with Autonomy and agentic scope: Agentic AI requires controls over actions and delegated authority as well as generated text. The present case instead concerns AI opacity and explanation.
Question 26
A lender's ensemble recommendation is highly accurate overall but borrowers cannot understand why their cases were flagged. The system owner is requesting a documented approval of this existing governance concern. Which safeguard most directly treats the described failure mode?
- Adopt suitable local explanations, decision records and a human challenge process without asserting that every feature is causal.
- Reassess risk classification, capacity for appeals and monitoring thresholds before scaling.
- Document the model, its inputs, predictions, decision points and accountable humans in the system inventory.
- Restrict tools by least privilege, define human approval boundaries and cap autonomous execution.
Explanation
Correct Answer
A:Adopt suitable local explanations, decision records and a human challenge process without asserting that every feature is causal. Correct. It directly reduces or governs the specific documented risk. Transparency and useful explainability must match the affected audience and impact, not simply display a score.
Incorrect Answers
B:Reassess risk classification, capacity for appeals and monitoring thresholds before scaling. Not selected. That action treats another control objective, not the specific gap described. It belongs with Scale-driven harm: Even a low individual error rate may create material harm when AI is applied at scale. The present case instead concerns AI opacity and explanation.
C:Document the model, its inputs, predictions, decision points and accountable humans in the system inventory. Not selected. That action treats another control objective, not the specific gap described. It belongs with AI system boundary: AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. The present case instead concerns AI opacity and explanation.
D:Restrict tools by least privilege, define human approval boundaries and cap autonomous execution. Not selected. That action treats another control objective, not the specific gap described. It belongs with Autonomy and agentic scope: Agentic AI requires controls over actions and delegated authority as well as generated text. The present case instead concerns AI opacity and explanation.
Question 27
A lender's ensemble recommendation is highly accurate overall but borrowers cannot understand why their cases were flagged. The system owner is requesting a documented approval of this existing governance concern. Which verification task should support the next approval decision?
- Attempt unauthorized and irreversible test actions and confirm the agent stops or requests approval.
- Trace test cases from model output to the human decision, including cases in which the output is ignored.
- Test whether affected users and reviewers can understand and contest representative decisions.
- Verify the fairness test improves relevant subgroup outcomes without exceeding approved data purposes.
Explanation
Correct Answer
C:Test whether affected users and reviewers can understand and contest representative decisions. Correct. It yields measurable evidence that the case-specific safeguard works. Transparency and useful explainability must match the affected audience and impact, not simply display a score.
Incorrect Answers
A:Attempt unauthorized and irreversible test actions and confirm the agent stops or requests approval. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Autonomy and agentic scope: Agentic AI requires controls over actions and delegated authority as well as generated text. The present case instead concerns AI opacity and explanation.
B:Trace test cases from model output to the human decision, including cases in which the output is ignored. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with AI system boundary: AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. The present case instead concerns AI opacity and explanation.
D:Verify the fairness test improves relevant subgroup outcomes without exceeding approved data purposes. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Responsible AI tradeoffs: Responsible AI principles interact; governance should record and justify risk tradeoffs rather than invoke a single slogan. The present case instead concerns AI opacity and explanation.
Question 28
A lender's ensemble recommendation is highly accurate overall but borrowers cannot understand why their cases were flagged. The system owner is requesting a documented approval of this existing governance concern. If the control fails, which immediate response most appropriately limits the documented risk?
- Pause the unexplained automated ranking path until system boundaries and oversight are documented.
- Suspend unsupported automated adverse recommendations and route cases for explainable review.
- Disable high-impact tool actions and revoke credentials if the agent exceeds its approved autonomy.
- Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning.
Explanation
Correct Answer
B:Suspend unsupported automated adverse recommendations and route cases for explainable review. Correct. It limits further reliance or exposure while the risk remains unresolved. Transparency and useful explainability must match the affected audience and impact, not simply display a score.
Incorrect Answers
A:Pause the unexplained automated ranking path until system boundaries and oversight are documented. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with AI system boundary: AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. The present case instead concerns AI opacity and explanation.
C:Disable high-impact tool actions and revoke credentials if the agent exceeds its approved autonomy. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with Autonomy and agentic scope: Agentic AI requires controls over actions and delegated authority as well as generated text. The present case instead concerns AI opacity and explanation.
D:Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning. Not selected. This action relates to AI system boundary. AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. For the contain decision, it does not resolve the specifically described issue of AI opacity and explanation in A lender's ensemble recommendation is highly accurate overall but borrowers cannot understand why their cases were flagged. The system owner is requesting a documented approval of this existing governance concern.
Question 29
A lender's ensemble recommendation is highly accurate overall but borrowers cannot understand why their cases were flagged. A separate assurance review finds the concern is relevant to an expanded user group. Which initial evidence check most directly addresses the governance concern?
- Identify the competing fairness, privacy, utility and lawful-processing objectives in the specific use case.
- Estimate aggregate impact using exposure volume, severity and unequal subgroup effects.
- Inspect calibration curves and subgroup error rates, not just the displayed confidence percentage.
- Identify whether explanations describe case-specific factors, system limits and meaningful review routes.
Explanation
Correct Answer
D:Identify whether explanations describe case-specific factors, system limits and meaningful review routes. Correct. It establishes the relevant factual risk boundary before implementation. Transparency and useful explainability must match the affected audience and impact, not simply display a score.
Incorrect Answers
A:Identify the competing fairness, privacy, utility and lawful-processing objectives in the specific use case. Not selected. That check primarily diagnoses a different governance issue. It belongs with Responsible AI tradeoffs: Responsible AI principles interact; governance should record and justify risk tradeoffs rather than invoke a single slogan. The present case instead concerns AI opacity and explanation.
B:Estimate aggregate impact using exposure volume, severity and unequal subgroup effects. Not selected. That check primarily diagnoses a different governance issue. It belongs with Scale-driven harm: Even a low individual error rate may create material harm when AI is applied at scale. The present case instead concerns AI opacity and explanation.
C:Inspect calibration curves and subgroup error rates, not just the displayed confidence percentage. Not selected. That check primarily diagnoses a different governance issue. It belongs with Probabilistic uncertainty: A probabilistic output is not a deterministic guarantee, especially in high-impact applications. The present case instead concerns AI opacity and explanation.
Question 30
A lender's ensemble recommendation is highly accurate overall but borrowers cannot understand why their cases were flagged. A separate assurance review finds the concern is relevant to an expanded user group. Which specific implementation best addresses the identified gap?
- Adopt suitable local explanations, decision records and a human challenge process without asserting that every feature is causal.
- Set confidence thresholds and an escalation pathway for uncertainty, with clinician review for consequential cases.
- Select a documented proportionate mitigation that tests fairness while minimizing sensitive-data access and securing any justified processing.
- Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning.
Explanation
Correct Answer
A:Adopt suitable local explanations, decision records and a human challenge process without asserting that every feature is causal. Correct. It directly reduces or governs the specific documented risk. Transparency and useful explainability must match the affected audience and impact, not simply display a score.
Incorrect Answers
B:Set confidence thresholds and an escalation pathway for uncertainty, with clinician review for consequential cases. Not selected. That action treats another control objective, not the specific gap described. It belongs with Probabilistic uncertainty: A probabilistic output is not a deterministic guarantee, especially in high-impact applications. The present case instead concerns AI opacity and explanation.
C:Select a documented proportionate mitigation that tests fairness while minimizing sensitive-data access and securing any justified processing. Not selected. That action treats another control objective, not the specific gap described. It belongs with Responsible AI tradeoffs: Responsible AI principles interact; governance should record and justify risk tradeoffs rather than invoke a single slogan. The present case instead concerns AI opacity and explanation.
D:Identify which component infers patterns from data and where outputs affect applicants, rather than treating every rule as machine learning. Not selected. This action relates to AI system boundary. AI governance starts with a precise inventory of what the AI system does and how its outputs affect decisions. For the control decision, it does not resolve the specifically described issue of AI opacity and explanation in A lender's ensemble recommendation is highly accurate overall but borrowers cannot understand why their cases were flagged. A separate assurance review finds the concern is relevant to an expanded user group.