ISACA Advanced in AI Audit AAIA Practice Test 11
Question 100
A generative service stores raw customer prompts indefinitely even though the approved purpose is short-term support. From an AI audit perspective, which corrective action is the MOST direct response to the weakness described, rather than a useful control for some different AI risk?
- Enforce least-privilege project entitlements and scoped deidentified training views
- Sanitize and segregate prompt traces with restricted access and retention
- Apply category-specific prompt retention limits, redaction and verified deletion paths
- Evaluate memorization, membership inference and release utility before sharing
Explanation
Correct Answer
C:Correct. Apply category-specific prompt retention limits, redaction and verified deletion paths is aimed at the actual failure: retention defaults violate purpose limitation and approved data minimization. The audit should subsequently obtain configured ttl, deletion test logs and contractual retention commitments to test execution, not merely the existence of a policy.
Incorrect Answers
A:“Enforce least-privilege project entitlements and scoped deidentified training views” is a control for dataset access is broad and lacks purpose-based authorization, but the audit finding here is retention defaults violate purpose limitation and approved data minimization. The relevant measure is “Apply category-specific prompt retention limits, redaction and verified deletion paths”.
B:“Sanitize and segregate prompt traces with restricted access and retention” is a control for observability logs bypass personal-data classification and handling controls, but the audit finding here is retention defaults violate purpose limitation and approved data minimization. The relevant measure is “Apply category-specific prompt retention limits, redaction and verified deletion paths”.
D:“Evaluate memorization, membership inference and release utility before sharing” is a control for synthetic data was assumed private without membership-inference or disclosure tests, but the audit finding here is retention defaults violate purpose limitation and approved data minimization. The relevant measure is “Apply category-specific prompt retention limits, redaction and verified deletion paths”.
Question 101
In an audit of Privacy and Data Governance Programs, managers say this measure is live: Apply category-specific prompt retention limits, redaction and verified deletion paths. They claim it corrected a material audit gap. What direct operating evidence would allow the auditor to substantiate that particular claim across a defensible sample?
- Access grants, denied cross-project query tests and periodic privilege recertifications
- Configured TTL, deletion test logs and contractual retention commitments
- Log pipeline samples, redaction tests and access reviews
- Attack evaluation reports, privacy thresholds and restricted release approvals
Explanation
Correct Answer
B:Correct. The proposed records—configured ttl, deletion test logs and contractual retention commitments—are directly tied to the asserted safeguard and allow its actual operation to be corroborated. The gap being tested is retention defaults violate purpose limitation and approved data minimization.
Incorrect Answers
A:“Access grants, denied cross-project query tests and periodic privilege recertifications” would test dataset access is broad and lacks purpose-based authorization, not the present safeguard against retention defaults violate purpose limitation and approved data minimization. The expected audit evidence is “Configured TTL, deletion test logs and contractual retention commitments”.
C:“Log pipeline samples, redaction tests and access reviews” would test observability logs bypass personal-data classification and handling controls, not the present safeguard against retention defaults violate purpose limitation and approved data minimization. The expected audit evidence is “Configured TTL, deletion test logs and contractual retention commitments”.
D:“Attack evaluation reports, privacy thresholds and restricted release approvals” would test synthetic data was assumed private without membership-inference or disclosure tests, not the present safeguard against retention defaults violate purpose limitation and approved data minimization. The expected audit evidence is “Configured TTL, deletion test logs and contractual retention commitments”.
Question 102
In a review of Privacy and Data Governance Programs, investigators cannot establish the expected control proof: configured ttl, deletion test logs and contractual retention commitments. The underlying event was that a generative service stores raw customer prompts indefinitely even though the approved purpose is short-term support. Which control weakness is the MOST likely cause requiring focused audit work?
- Retention defaults violate purpose limitation and approved data minimization
- Dataset access is broad and lacks purpose-based authorization
- Observability logs bypass personal-data classification and handling controls
- Synthetic data was assumed private without membership-inference or disclosure tests
Explanation
Correct Answer
A:Correct. The observed condition specifically indicates retention defaults violate purpose limitation and approved data minimization. A meaningful audit response is to verify this diagnosis and then recommend apply category-specific prompt retention limits, redaction and verified deletion paths.
Incorrect Answers
B:“Dataset access is broad and lacks purpose-based authorization” concerns a separate audit weakness, not the missing assurance described by retention defaults violate purpose limitation and approved data minimization. The correct cause is “Retention defaults violate purpose limitation and approved data minimization”.
C:“Observability logs bypass personal-data classification and handling controls” concerns a separate audit weakness, not the missing assurance described by retention defaults violate purpose limitation and approved data minimization. The correct cause is “Retention defaults violate purpose limitation and approved data minimization”.
D:“Synthetic data was assumed private without membership-inference or disclosure tests” concerns a separate audit weakness, not the missing assurance described by retention defaults violate purpose limitation and approved data minimization. The correct cause is “Retention defaults violate purpose limitation and approved data minimization”.
Question 103
Training data was purchased from a broker but the vendor cannot demonstrate collection rights. From an AI audit perspective, which corrective action is the MOST direct response to the weakness described, rather than a useful control for some different AI risk?
- Conduct purpose/compatibility and privacy impact review before reuse
- Extend deletion propagation to indexes, caches, fine-tune stores and backups as applicable
- Test extraction risk and restrict sensitive content in model training and output handling
- Require documented lineage, rights basis and permitted-use checks before ingestion
Explanation
Correct Answer
D:Correct. Require documented lineage, rights basis and permitted-use checks before ingestion is aimed at the actual failure: data provenance and lawful-use assurance are not established. The audit should subsequently obtain data source records, licensing terms and approval evidence for each training corpus to test execution, not merely the existence of a policy.
Incorrect Answers
A:“Conduct purpose/compatibility and privacy impact review before reuse” is a control for secondary AI processing is not evaluated against the original lawful purpose, but the audit finding here is data provenance and lawful-use assurance are not established. The relevant measure is “Require documented lineage, rights basis and permitted-use checks before ingestion”.
B:“Extend deletion propagation to indexes, caches, fine-tune stores and backups as applicable” is a control for subject-rights workflow excludes derived AI data stores and vector copies, but the audit finding here is data provenance and lawful-use assurance are not established. The relevant measure is “Require documented lineage, rights basis and permitted-use checks before ingestion”.
C:“Test extraction risk and restrict sensitive content in model training and output handling” is a control for training memorization and output disclosure controls were unassessed, but the audit finding here is data provenance and lawful-use assurance are not established. The relevant measure is “Require documented lineage, rights basis and permitted-use checks before ingestion”.
Question 104
In an audit of Privacy and Data Governance Programs, managers say this measure is live: Require documented lineage, rights basis and permitted-use checks before ingestion. They claim it corrected a material audit gap. What direct operating evidence would allow the auditor to substantiate that particular claim across a defensible sample?
- Updated processing register, legal assessment and approved purpose mapping
- End-to-end deletion trace from source to vector index and exception ledger
- Data source records, licensing terms and approval evidence for each training corpus
- Memorization probe results, content filters and reviewed training-data exclusions
Explanation
Correct Answer
C:Correct. The proposed records—data source records, licensing terms and approval evidence for each training corpus—are directly tied to the asserted safeguard and allow its actual operation to be corroborated. The gap being tested is data provenance and lawful-use assurance are not established.
Incorrect Answers
A:“Updated processing register, legal assessment and approved purpose mapping” would test secondary AI processing is not evaluated against the original lawful purpose, not the present safeguard against data provenance and lawful-use assurance are not established. The expected audit evidence is “Data source records, licensing terms and approval evidence for each training corpus”.
B:“End-to-end deletion trace from source to vector index and exception ledger” would test subject-rights workflow excludes derived AI data stores and vector copies, not the present safeguard against data provenance and lawful-use assurance are not established. The expected audit evidence is “Data source records, licensing terms and approval evidence for each training corpus”.
D:“Memorization probe results, content filters and reviewed training-data exclusions” would test training memorization and output disclosure controls were unassessed, not the present safeguard against data provenance and lawful-use assurance are not established. The expected audit evidence is “Data source records, licensing terms and approval evidence for each training corpus”.
Question 105
In a review of Privacy and Data Governance Programs, investigators cannot establish the expected control proof: data source records, licensing terms and approval evidence for each training corpus. The underlying event was that training data was purchased from a broker but the vendor cannot demonstrate collection rights. Which control weakness is the MOST likely cause requiring focused audit work?
- Secondary AI processing is not evaluated against the original lawful purpose
- Data provenance and lawful-use assurance are not established
- Subject-rights workflow excludes derived AI data stores and vector copies
- Training memorization and output disclosure controls were unassessed
Explanation
Correct Answer
B:Correct. The observed condition specifically indicates data provenance and lawful-use assurance are not established. A meaningful audit response is to verify this diagnosis and then recommend require documented lineage, rights basis and permitted-use checks before ingestion.
Incorrect Answers
A:“Secondary AI processing is not evaluated against the original lawful purpose” concerns a separate audit weakness, not the missing assurance described by data provenance and lawful-use assurance are not established. The correct cause is “Data provenance and lawful-use assurance are not established”.
C:“Subject-rights workflow excludes derived AI data stores and vector copies” concerns a separate audit weakness, not the missing assurance described by data provenance and lawful-use assurance are not established. The correct cause is “Data provenance and lawful-use assurance are not established”.
D:“Training memorization and output disclosure controls were unassessed” concerns a separate audit weakness, not the missing assurance described by data provenance and lawful-use assurance are not established. The correct cause is “Data provenance and lawful-use assurance are not established”.
Question 106
A hospital's model-building team can query every patient's history regardless of assigned project. From an AI audit perspective, which corrective action is the MOST direct response to the weakness described, rather than a useful control for some different AI risk?
- Enforce least-privilege project entitlements and scoped deidentified training views
- Perform linkage attacks and set release safeguards appropriate to residual identifiability
- Capture versioned transformation logic, imputation decisions and field-level lineage
- Apply category-specific prompt retention limits, redaction and verified deletion paths
Explanation
Correct Answer
A:Correct. Enforce least-privilege project entitlements and scoped deidentified training views is aimed at the actual failure: dataset access is broad and lacks purpose-based authorization. The audit should subsequently obtain access grants, denied cross-project query tests and periodic privilege recertifications to test execution, not merely the existence of a policy.
Incorrect Answers
B:“Perform linkage attacks and set release safeguards appropriate to residual identifiability” is a control for reidentification risk was not tested for the actual release context, but the audit finding here is dataset access is broad and lacks purpose-based authorization. The relevant measure is “Enforce least-privilege project entitlements and scoped deidentified training views”.
C:“Capture versioned transformation logic, imputation decisions and field-level lineage” is a control for lineage controls omit preprocessing operations essential to auditability, but the audit finding here is dataset access is broad and lacks purpose-based authorization. The relevant measure is “Enforce least-privilege project entitlements and scoped deidentified training views”.
D:“Apply category-specific prompt retention limits, redaction and verified deletion paths” is a control for retention defaults violate purpose limitation and approved data minimization, but the audit finding here is dataset access is broad and lacks purpose-based authorization. The relevant measure is “Enforce least-privilege project entitlements and scoped deidentified training views”.
Question 107
In an audit of Privacy and Data Governance Programs, managers say this measure is live: Enforce least-privilege project entitlements and scoped deidentified training views. They claim it corrected a material audit gap. What direct operating evidence would allow the auditor to substantiate that particular claim across a defensible sample?
- Adversarial reidentification results, suppression thresholds and release decision
- Reproducible pipeline manifests and source-to-feature trace samples
- Configured TTL, deletion test logs and contractual retention commitments
- Access grants, denied cross-project query tests and periodic privilege recertifications
Explanation
Correct Answer
D:Correct. The proposed records—access grants, denied cross-project query tests and periodic privilege recertifications—are directly tied to the asserted safeguard and allow its actual operation to be corroborated. The gap being tested is dataset access is broad and lacks purpose-based authorization.
Incorrect Answers
A:“Adversarial reidentification results, suppression thresholds and release decision” would test reidentification risk was not tested for the actual release context, not the present safeguard against dataset access is broad and lacks purpose-based authorization. The expected audit evidence is “Access grants, denied cross-project query tests and periodic privilege recertifications”.
B:“Reproducible pipeline manifests and source-to-feature trace samples” would test lineage controls omit preprocessing operations essential to auditability, not the present safeguard against dataset access is broad and lacks purpose-based authorization. The expected audit evidence is “Access grants, denied cross-project query tests and periodic privilege recertifications”.
C:“Configured TTL, deletion test logs and contractual retention commitments” would test retention defaults violate purpose limitation and approved data minimization, not the present safeguard against dataset access is broad and lacks purpose-based authorization. The expected audit evidence is “Access grants, denied cross-project query tests and periodic privilege recertifications”.
Question 108
In a review of Privacy and Data Governance Programs, investigators cannot establish the expected control proof: access grants, denied cross-project query tests and periodic privilege recertifications. The underlying event was that a hospital's model-building team can query every patient's history regardless of assigned project. Which control weakness is the MOST likely cause requiring focused audit work?
- Reidentification risk was not tested for the actual release context
- Lineage controls omit preprocessing operations essential to auditability
- Dataset access is broad and lacks purpose-based authorization
- Retention defaults violate purpose limitation and approved data minimization
Explanation
Correct Answer
C:Correct. The observed condition specifically indicates dataset access is broad and lacks purpose-based authorization. A meaningful audit response is to verify this diagnosis and then recommend enforce least-privilege project entitlements and scoped deidentified training views.
Incorrect Answers
A:“Reidentification risk was not tested for the actual release context” concerns a separate audit weakness, not the missing assurance described by dataset access is broad and lacks purpose-based authorization. The correct cause is “Dataset access is broad and lacks purpose-based authorization”.
B:“Lineage controls omit preprocessing operations essential to auditability” concerns a separate audit weakness, not the missing assurance described by dataset access is broad and lacks purpose-based authorization. The correct cause is “Dataset access is broad and lacks purpose-based authorization”.
D:“Retention defaults violate purpose limitation and approved data minimization” concerns a separate audit weakness, not the missing assurance described by dataset access is broad and lacks purpose-based authorization. The correct cause is “Dataset access is broad and lacks purpose-based authorization”.
Question 109
Customer preference data collected for delivery scheduling is reused for credit scoring without a new assessment. From an AI audit perspective, which corrective action is the MOST direct response to the weakness described, rather than a useful control for some different AI risk?
- Sanitize and segregate prompt traces with restricted access and retention
- Conduct purpose/compatibility and privacy impact review before reuse
- Inventory processing locations, subprocessors and cross-border transfer safeguards
- Require documented lineage, rights basis and permitted-use checks before ingestion
Explanation
Correct Answer
B:Correct. Conduct purpose/compatibility and privacy impact review before reuse is aimed at the actual failure: secondary ai processing is not evaluated against the original lawful purpose. The audit should subsequently obtain updated processing register, legal assessment and approved purpose mapping to test execution, not merely the existence of a policy.
Incorrect Answers
A:“Sanitize and segregate prompt traces with restricted access and retention” is a control for observability logs bypass personal-data classification and handling controls, but the audit finding here is secondary AI processing is not evaluated against the original lawful purpose. The relevant measure is “Conduct purpose/compatibility and privacy impact review before reuse”.
C:“Inventory processing locations, subprocessors and cross-border transfer safeguards” is a control for aI data transfers are missing location and processor transparency, but the audit finding here is secondary AI processing is not evaluated against the original lawful purpose. The relevant measure is “Conduct purpose/compatibility and privacy impact review before reuse”.
D:“Require documented lineage, rights basis and permitted-use checks before ingestion” is a control for data provenance and lawful-use assurance are not established, but the audit finding here is secondary AI processing is not evaluated against the original lawful purpose. The relevant measure is “Conduct purpose/compatibility and privacy impact review before reuse”.
Question 110
In an audit of Privacy and Data Governance Programs, managers say this measure is live: Conduct purpose/compatibility and privacy impact review before reuse. They claim it corrected a material audit gap. What direct operating evidence would allow the auditor to substantiate that particular claim across a defensible sample?
- Updated processing register, legal assessment and approved purpose mapping
- Log pipeline samples, redaction tests and access reviews
- Verified processor list, hosting configuration and applicable transfer assessments
- Data source records, licensing terms and approval evidence for each training corpus
Explanation
Correct Answer
A:Correct. The proposed records—updated processing register, legal assessment and approved purpose mapping—are directly tied to the asserted safeguard and allow its actual operation to be corroborated. The gap being tested is secondary ai processing is not evaluated against the original lawful purpose.
Incorrect Answers
B:“Log pipeline samples, redaction tests and access reviews” would test observability logs bypass personal-data classification and handling controls, not the present safeguard against secondary AI processing is not evaluated against the original lawful purpose. The expected audit evidence is “Updated processing register, legal assessment and approved purpose mapping”.
C:“Verified processor list, hosting configuration and applicable transfer assessments” would test aI data transfers are missing location and processor transparency, not the present safeguard against secondary AI processing is not evaluated against the original lawful purpose. The expected audit evidence is “Updated processing register, legal assessment and approved purpose mapping”.
D:“Data source records, licensing terms and approval evidence for each training corpus” would test data provenance and lawful-use assurance are not established, not the present safeguard against secondary AI processing is not evaluated against the original lawful purpose. The expected audit evidence is “Updated processing register, legal assessment and approved purpose mapping”.