Practice Exams:

IAPP AIGP Dumps

EXAM: Artificial Intelligence Governance Professional
Verified By Experts
Premium Bundle
$40.00
$59.97 $99.97

AIGP Premium Bundle


  • Premium File 222 Questions & Answers
  • Training Course 124 Lectures
  • Study Guide 566 Pages

IAPP AIGP Practice Test 22

Question 210

Choose the intervention and validation that fit this business problem: A health startup stores voice biometrics for account access and assumes pseudonyms remove privacy duties.

  1. Action: Identify and document the applicable lawful basis before reusing personal information for model development. Evidence: Check the data-use assessment for jurisdiction, purpose, applicable lawful basis and approval before ingestion.
  2. Action: Escalate biometric or other specially protected personal data for a jurisdiction-specific legal and security review. Evidence: Confirm the applicable category, explicit necessity assessment, protective controls and approved processing conditions.
  3. Action: Remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient. Evidence: Compare approved feature requirements with actual datasets and test a reduced-data version for utility.
  4. Action: Review AI-related consumer claims for accuracy, substantiation and misleading omissions before marketing or use. Evidence: Compare published promises, disclaimers and actual tested limitations before approval.

Explanation

Correct Answer

B:

For the action and verification, escalate biometric or other specially protected personal data for a jurisdiction-specific legal and security review. Confirm the applicable category, explicit necessity assessment, protective controls and approved processing conditions. Together, they implement and test the intended safeguard. Sensitive data can trigger heightened restrictions and requires particular attention to necessity, safeguards and permitted uses

Incorrect Answers

A:

For the action and verification, this pair supports another governance intervention: Identify and document the applicable lawful basis before reusing personal information for model development. For the actual case, the required action is: Escalate biometric or other specially protected personal data for a jurisdiction-specific legal and security review. The supporting evidence is: Confirm the applicable category, explicit necessity assessment, protective controls and approved processing conditions

C:

This pair supports another governance intervention: Remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient. For the actual case, the required action is: Escalate biometric or other specially protected personal data for a jurisdiction-specific legal and security review. The supporting evidence is: Confirm the applicable category, explicit necessity assessment, protective controls and approved processing conditions.

D:

This pair supports another governance intervention: Review AI-related consumer claims for accuracy, substantiation and misleading omissions before marketing or use. For the actual case, the required action is: Escalate biometric or other specially protected personal data for a jurisdiction-specific legal and security review. The supporting evidence is: Confirm the applicable category, explicit necessity assessment, protective controls and approved processing conditions.

Question 211

An AI project team faces this situation: A bank deploys automated account flags and lacks a process for customers to challenge incorrect personal data. What should the team do first to manage the central issue?

  1. Remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient.
  2. Test whether a proposed new model purpose is compatible with the reason personal information was originally obtained.
  3. Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards.
  4. Keep a current evidence map of relevant requirements, assessments and decision records for regulator or conformity review.

Explanation

Correct Answer

C:

Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable. The practical acceptance evidence is: test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

Incorrect Answers

A:

The alternative—remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient—is not the most direct response to this case. The matching action is to design a process to identify relevant data subject requests and assess applicable automated-decision safeguards.

B:

The alternative—test whether a proposed new model purpose is compatible with the reason personal information was originally obtained—is not the most direct response to this case. The matching action is to design a process to identify relevant data subject requests and assess applicable automated-decision safeguards.

D:

The alternative—keep a current evidence map of relevant requirements, assessments and decision records for regulator or conformity review—is not the most direct response to this case. The matching action is to design a process to identify relevant data subject requests and assess applicable automated-decision safeguards.

Question 212

A team reviewing the following situation asks for the most relevant governance principle: A bank deploys automated account flags and lacks a process for customers to challenge incorrect personal data. Which principle best explains the required control?

  1. Sensitive data can trigger heightened restrictions and requires particular attention to necessity, safeguards and permitted uses.
  2. Having technical access to records does not, by itself, authorize a new AI training purpose.
  3. Public visibility of material does not establish blanket rights to copy or use it for every machine-learning purpose.
  4. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

Explanation

Correct Answer

D:

AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable. That is why the case calls for this action: design a process to identify relevant data subject requests and assess applicable automated-decision safeguards.

Incorrect Answers

A:

That explains why a team might escalate biometric or other specially protected personal data for a jurisdiction-specific legal and security review, but it does not justify the decision required by the case. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

B:

That explains why a team might identify and document the applicable lawful basis before reusing personal information for model development, but it does not justify the decision required by the case. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

C:

That explains why a team might document rights and restrictions for copyrighted or licensed sources before training or adapting a model, but it does not justify the decision required by the case. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

Question 213

A bank deploys automated account flags and lacks a process for customers to challenge incorrect personal data. Which recorded test or observation most directly demonstrates an effective response?

  1. Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.
  2. Record the original notice and collection purpose alongside the proposed use and compatibility determination.
  3. Compare approved feature requirements with actual datasets and test a reduced-data version for utility.
  4. Record the affected people, purpose, jurisdiction, current legal classification and legal disposition.

Explanation

Correct Answer

A:

Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs. This verifies the necessary safeguard rather than merely showing that an AI project was completed. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

Incorrect Answers

B:

Record the original notice and collection purpose alongside the proposed use and compatibility determination would test a different outcome. The directly relevant evidence is to test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

C:

Compare approved feature requirements with actual datasets and test a reduced-data version for utility would test a different outcome. The directly relevant evidence is to test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

D:

Record the affected people, purpose, jurisdiction, current legal classification and legal disposition would test a different outcome. The directly relevant evidence is to test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

Question 214

For this case—A bank deploys automated account flags and lacks a process for customers to challenge incorrect personal data—staff choose the shortcut: Tell rejected applicants that model decisions can never be reviewed because the system is automated. Which consequence best explains why this is inadequate?

  1. Identifiable or sensitive characteristics may still be exposed or processed unlawfully.
  2. People may lose legally required recourse or receive no usable explanation of a consequential outcome.
  3. People's data may support decisions they were never reasonably told to expect.
  4. The organization may be unable to show compliance or respond effectively to a formal inquiry.

Explanation

Correct Answer

B:

People may lose legally required recourse or receive no usable explanation of a consequential outcome. The shortcut overlooks the central safeguard: design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

Incorrect Answers

A:

This consequence follows more directly from another shortcut: Treat face embeddings as ordinary anonymous telemetry simply because names are removed. The shortcut actually described would instead create this risk: People may lose legally required recourse or receive no usable explanation of a consequential outcome.

C:

This consequence follows more directly from another shortcut: Declare every internal AI experiment a permitted extension of the original collection purpose. The shortcut actually described would instead create this risk: People may lose legally required recourse or receive no usable explanation of a consequential outcome.

D:

This consequence follows more directly from another shortcut: Wait for a regulator to ask questions before creating system documentation and records. The shortcut actually described would instead create this risk: People may lose legally required recourse or receive no usable explanation of a consequential outcome.

Question 215

The governance team needs both a decision and a way to verify it for this case: A bank deploys automated account flags and lacks a process for customers to challenge incorrect personal data. Which pair is best?

  1. Action: Remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient. Evidence: Compare approved feature requirements with actual datasets and test a reduced-data version for utility.
  2. Action: Test whether a proposed new model purpose is compatible with the reason personal information was originally obtained. Evidence: Record the original notice and collection purpose alongside the proposed use and compatibility determination.
  3. Action: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. Evidence: Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.
  4. Action: Include foreseeable AI failure modes, product interfaces and remedial controls in the relevant safety/legal assessment. Evidence: Review documented hazards, foreseeable misuse, safety testing and corrective action ownership.

Explanation

Correct Answer

C:

Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs. Together, they implement and test the intended safeguard. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

Incorrect Answers

A:

This pair supports another governance intervention: Remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient. For the actual case, the required action is: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. The supporting evidence is: Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

B:

This pair supports another governance intervention: Test whether a proposed new model purpose is compatible with the reason personal information was originally obtained. For the actual case, the required action is: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. The supporting evidence is: Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

D:

This pair supports another governance intervention: Include foreseeable AI failure modes, product interfaces and remedial controls in the relevant safety/legal assessment. For the actual case, the required action is: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. The supporting evidence is: Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

Question 216

An AI project team faces this situation: A recruitment platform provides rejection scores without checking obligations for human review and candidate requests. What should the team do first to manage the central issue?

  1. Identify and document the applicable lawful basis before reusing personal information for model development.
  2. Remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient.
  3. Review AI-related consumer claims for accuracy, substantiation and misleading omissions before marketing or use.
  4. Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards.

Explanation

Correct Answer

D:

For the decision, design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable. The practical acceptance evidence is: test sample requests for access, correction and review against the governing jurisdictional rules and operational logs

Incorrect Answers

A:

This would help address a separate governance task: Identify and document the applicable lawful basis before reusing personal information for model development. For this case, however, aI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable. The more direct step is: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards.

B:

For the decision, the alternative—remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient—is not the most direct response to this case. The matching action is to design a process to identify relevant data subject requests and assess applicable automated-decision safeguards

C:

This would help address a separate governance task: Review AI-related consumer claims for accuracy, substantiation and misleading omissions before marketing or use. For this case, however, aI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable. The more direct step is: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards.

Question 217

A recruitment platform provides rejection scores without checking obligations for human review and candidate requests. Why is the recommended governance approach appropriate?

  1. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.
  2. Having technical access to records does not, by itself, authorize a new AI training purpose.
  3. Reusing information for an unrelated objective can create a new privacy obligation despite good technical safeguards.
  4. Deceptive descriptions of model capabilities can create legal risk even when the underlying product functions.

Explanation

Correct Answer

A:

For the technical explanation, aI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable. That is why the case calls for this action: design a process to identify relevant data subject requests and assess applicable automated-decision safeguards

Incorrect Answers

B:

For the technical explanation, that explains why a team might identify and document the applicable lawful basis before reusing personal information for model development, but it does not justify the decision required by the case. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable

C:

That explains why a team might test whether a proposed new model purpose is compatible with the reason personal information was originally obtained, but it does not justify the decision required by the case. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

D:

That explains why a team might review AI-related consumer claims for accuracy, substantiation and misleading omissions before marketing or use, but it does not justify the decision required by the case. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable.

Question 218

A recruitment platform provides rejection scores without checking obligations for human review and candidate requests. Which recorded test or observation most directly demonstrates an effective response?

  1. Confirm the applicable category, explicit necessity assessment, protective controls and approved processing conditions.
  2. Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.
  3. Compare approved feature requirements with actual datasets and test a reduced-data version for utility.
  4. Inspect governance roles, risk tolerance and communication processes that support subsequent risk work.

Explanation

Correct Answer

B:

For the verification, test sample requests for access, correction and review against the governing jurisdictional rules and operational logs. This verifies the necessary safeguard rather than merely showing that an AI project was completed. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable

Incorrect Answers

A:

Confirm the applicable category, explicit necessity assessment, protective controls and approved processing conditions would test a different outcome. The directly relevant evidence is to test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

C:

For the verification, compare approved feature requirements with actual datasets and test a reduced-data version for utility would test a different outcome. The directly relevant evidence is to test sample requests for access, correction and review against the governing jurisdictional rules and operational logs

D:

Inspect governance roles, risk tolerance and communication processes that support subsequent risk work would test a different outcome. The directly relevant evidence is to test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

Question 219

A team considers this AI use: A recruitment platform provides rejection scores without checking obligations for human review and candidate requests. Rather than applying the relevant control, it will tell rejected applicants that model decisions can never be reviewed because the system is automated. What can go wrong?

  1. Unneeded sensitive fields can leak or influence outcomes while adding no justified benefit.
  2. People's data may support decisions they were never reasonably told to expect.
  3. People may lose legally required recourse or receive no usable explanation of a consequential outcome.
  4. The organization may build a use that cannot lawfully operate in its target context.

Explanation

Correct Answer

C:

For the risk comparison, people may lose legally required recourse or receive no usable explanation of a consequential outcome. The shortcut overlooks the central safeguard: design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable

Incorrect Answers

A:

This consequence follows more directly from another shortcut: Feed the entire customer database into the model because storage and compute are available. The shortcut actually described would instead create this risk: People may lose legally required recourse or receive no usable explanation of a consequential outcome.

B:

For the risk comparison, this consequence follows more directly from another shortcut: Declare every internal AI experiment a permitted extension of the original collection purpose. The shortcut actually described would instead create this risk: People may lose legally required recourse or receive no usable explanation of a consequential outcome

D:

This consequence follows more directly from another shortcut: Proceed with a sensitive behavior-profiling use case first and check prohibitions only after launch. The shortcut actually described would instead create this risk: People may lose legally required recourse or receive no usable explanation of a consequential outcome.

Question 220

Choose the intervention and validation that fit this business problem: A recruitment platform provides rejection scores without checking obligations for human review and candidate requests.

  1. Action: Remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient. Evidence: Compare approved feature requirements with actual datasets and test a reduced-data version for utility.
  2. Action: Identify and document the applicable lawful basis before reusing personal information for model development. Evidence: Check the data-use assessment for jurisdiction, purpose, applicable lawful basis and approval before ingestion.
  3. Action: Select ISO/IEC 42001 for AI management-system governance, ISO/IEC 22989 for terminology and concepts, and ISO/IEC 42005 for AI-system impact assessment guidance. Evidence: Verify that the chosen reference matches the control objective and that the project cites the correct standard.
  4. Action: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. Evidence: Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

Explanation

Correct Answer

D:

For the action and verification, design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs. Together, they implement and test the intended safeguard. AI-supported decisions do not automatically remove individuals' legal rights to information, contestation or appropriate review where applicable

Incorrect Answers

A:

For the action and verification, this pair supports another governance intervention: Remove unnecessary identifiers and features before training and inference, while testing whether utility remains sufficient. For the actual case, the required action is: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. The supporting evidence is: Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs

B:

This pair supports another governance intervention: Identify and document the applicable lawful basis before reusing personal information for model development. For the actual case, the required action is: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. The supporting evidence is: Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

C:

This pair supports another governance intervention: Select ISO/IEC 42001 for AI management-system governance, ISO/IEC 22989 for terminology and concepts, and ISO/IEC 42005 for AI-system impact assessment guidance. For the actual case, the required action is: Design a process to identify relevant data subject requests and assess applicable automated-decision safeguards. The supporting evidence is: Test sample requests for access, correction and review against the governing jurisdictional rules and operational logs.

Artificial Intelligence Governance Professional (AIGP) Practice Tests