Isaca AAIA Practice Test 82
Question 810
An independent AI auditor reviews this case: A service desk assistant retrieves a wiki page containing instructions to ignore the organization's security controls. Which TWO complementary elements should the auditor seek: an effective corrective control and the most probative independent evidence? Select TWO answers.
- Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable.
- Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
- Verify model provenance, storage integrity, executable dependencies and approved component changes.
- Treat retrieved text as untrusted data, separate instruction authority and constrain downstream tool access.
- Inspect index access, chunk hashes, ingest logs, tenant segregation and adversarial retrieval results.
Explanation
Correct Answers
B:The reason it is necessary is: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges. The corrective action is: Treat retrieved text as untrusted data, separate instruction authority and constrain downstream tool access.
D:The reason it is necessary is: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges. This response fits the risk associated with Indirect prompt injection through retrieved content: Treat retrieved text as untrusted data, separate instruction authority and constrain downstream tool access.
Incorrect Answers
A:This attractive shortcut does not resolve the Indirect prompt injection through retrieved content weakness: Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable. For this situation: A service desk assistant retrieves a wiki page containing instructions to ignore the organization's security controls.
C:This option concerns Foundation model dependency and artifact tampering, which is a related but different AI governance concern. The decisive principle here is: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges.
E:This option concerns Poisoned embedding and vector retrieval index, which is a related but different AI governance concern. The decisive principle here is: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges.
Question 811
A governance and assurance review uncovers the following: A service desk assistant retrieves a wiki page containing instructions to ignore the organization's security controls. What evidence should the auditor examine FIRST to substantiate or challenge the control owner’s claim? Select ONE best answer.
- Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
- Inspect index access, chunk hashes, ingest logs, tenant segregation and adversarial retrieval results.
- Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable.
- Inspect authentication, rate limits, output granularity, query anomalies and abuse response.
Explanation
Correct Answer
A:The reason it is necessary is: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges. For this situation: A service desk assistant retrieves a wiki page containing instructions to ignore the organization's security controls.
Incorrect Answers
B:This option concerns Poisoned embedding and vector retrieval index, which is a related but different AI governance concern. For this situation: A service desk assistant retrieves a wiki page containing instructions to ignore the organization's security controls.
C:This attractive shortcut does not resolve the Indirect prompt injection through retrieved content weakness: Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable. Treat retrieved text as untrusted data, separate instruction authority and constrain downstream tool access.
D:This option concerns AI model extraction and excessive query access, which is a related but different AI governance concern. The audit must corroborate: Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
Question 812
A risk-based audit engagement examines this operating condition: A service desk assistant retrieves a wiki page containing instructions to ignore the organization's security controls. After management addresses the issue, which result MOST convincingly demonstrates effective remediation? Select ONE best answer.
- Suspicious extraction patterns are detected and unauthorized systematic probing is constrained.
- An independent rerun can identify the exact source data and material transformation decisions for a sampled model version.
- Malicious document instructions cannot cause prohibited tool actions or confidential-data release.
- Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable.
Explanation
Correct Answer
C:The reason it is necessary is: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges. Independent audit evidence includes: Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
Incorrect Answers
A:This option concerns AI model extraction and excessive query access, which is a related but different AI governance concern. For this situation: A service desk assistant retrieves a wiki page containing instructions to ignore the organization's security controls.
B:This option concerns AI training dataset lineage and reproducibility, which is a related but different AI governance concern. For this situation: A service desk assistant retrieves a wiki page containing instructions to ignore the organization's security controls.
D:It misses the key control principle: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges.
Question 813
An audit committee asks for a defensible assessment of this situation: A customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets. What is the MOST appropriate risk-based audit recommendation that addresses the root control weakness? Select ONE best answer.
- Version prompt templates, permission boundaries and evaluation suites with staged approvals and rollback.
- Evaluate unintended inference from model outputs and minimize observable detail and query access.
- Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable.
- Treat retrieved text as untrusted data, separate instruction authority and constrain downstream tool access.
Explanation
Correct Answer
D:The reason it is necessary is: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges. In this case: A customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets.
Incorrect Answers
A:This option concerns Prompt template governance and injection-resilience changes, which is a related but different AI governance concern. The audit must corroborate: Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
B:This option concerns Model inversion and sensitive attribute leakage, which is a related but different AI governance concern. It does not establish control effectiveness for Indirect prompt injection through retrieved content.
C:This attractive shortcut does not resolve the Indirect prompt injection through retrieved content weakness: Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable. In this case: A customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets.
Question 814
During an AI controls assessment, the team learns that a customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets. Which principle BEST explains why the observed practice is a material governance or control concern? Select ONE best answer.
- Model weights, parsers and surrounding runtime code belong in the software and data supply-chain threat model.
- Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges.
- Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable.
- Model-generated text is untrusted as an authorization basis for real-world privileged actions.
Explanation
Correct Answer
B:The acceptable verified outcome is: Malicious document instructions cannot cause prohibited tool actions or confidential-data release. Independent audit evidence includes: Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
Incorrect Answers
A:This option concerns Foundation model dependency and artifact tampering, which is a related but different AI governance concern. The audit must corroborate: Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
C:This attractive shortcut does not resolve the Indirect prompt injection through retrieved content weakness: Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable. For this situation: A customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets.
D:This option concerns AI tool misuse and excessive agency, which is a related but different AI governance concern. The decisive principle here is: Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges.
Question 815
An independent AI auditor reviews this case: A customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets. What evidence should the auditor examine FIRST to substantiate or challenge the control owner’s claim? Select ONE best answer.
- Review label guidelines, inter-rater reliability, adjudication logs, sampled errors and retraining effects.
- Inspect dataset snapshots, feature pipeline commits, record counts, model registry links and training execution manifests.
- Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable.
- Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
Explanation
Correct Answer
D:The acceptable verified outcome is: Malicious document instructions cannot cause prohibited tool actions or confidential-data release.
Incorrect Answers
A:This option concerns Label quality and human annotator agreement, which is a related but different AI governance concern. It does not establish control effectiveness for Indirect prompt injection through retrieved content.
B:This option concerns AI training dataset lineage and reproducibility, which is a related but different AI governance concern. The outcome to verify is: Malicious document instructions cannot cause prohibited tool actions or confidential-data release.
C:For this situation: A customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets.
Question 816
A governance and assurance review uncovers the following: A customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets. After management addresses the issue, which result MOST convincingly demonstrates effective remediation? Select ONE best answer.
- Only approved corpus changes influence outputs, and unexpected ranking changes trigger investigation.
- Assume any instruction embedded in a retrieved paragraph is authorized because the document is searchable.
- Malicious document instructions cannot cause prohibited tool actions or confidential-data release.
- Authorized prompt versions perform as specified and fail closed on tested malicious context.
Explanation
Correct Answer
C:This is an observable validation result for Indirect prompt injection through retrieved content, not just management assurance: Malicious document instructions cannot cause prohibited tool actions or confidential-data release.
Incorrect Answers
A:This option concerns Poisoned embedding and vector retrieval index, which is a related but different AI governance concern. The audit must corroborate: Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
B:The auditor should instead examine: Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
D:This option concerns Prompt template governance and injection-resilience changes, which is a related but different AI governance concern. For this situation: A customer chatbot reads a malicious webpage that poses as a system administrator and requests hidden secrets.
Question 817
A risk-based audit engagement examines this operating condition: A competitor repeatedly queries a proprietary prediction API to recreate its behavior and pricing decisions. What is the MOST appropriate risk-based audit recommendation that addresses the root control weakness? Select ONE best answer.
- Scope tools to specific identities and approved operations with pre-execution authorization and transaction caps.
- Authenticate retrieval corpus updates, isolate tenants and test ranking anomalies and source integrity.
- Rate-limit and monitor queries, restrict unnecessary output detail and evaluate extraction risks proportionate to value.
- Assume model theft requires downloading raw parameters and leave public prediction queries unrestricted.
Explanation
Correct Answer
C:The reason it is necessary is: Repeated API observations can reveal model behavior or sensitive correlations even without access to the weight files. Independent audit evidence includes: Inspect authentication, rate limits, output granularity, query anomalies and abuse response.
Incorrect Answers
A:This option concerns AI tool misuse and excessive agency, which is a related but different AI governance concern. It does not establish control effectiveness for AI model extraction and excessive query access.
B:This option concerns Poisoned embedding and vector retrieval index, which is a related but different AI governance concern. It does not establish control effectiveness for AI model extraction and excessive query access.
D:This attractive shortcut does not resolve the AI model extraction and excessive query access weakness: Assume model theft requires downloading raw parameters and leave public prediction queries unrestricted.
Question 818
An audit committee asks for a defensible assessment of this situation: A competitor repeatedly queries a proprietary prediction API to recreate its behavior and pricing decisions. Which principle BEST explains why the observed practice is a material governance or control concern? Select ONE best answer.
- Assume model theft requires downloading raw parameters and leave public prediction queries unrestricted.
- An AI incident ends only after decision impacts, service restoration and recurrence risk are addressed.
- Untrusted retrieved documents can manipulate model instructions but must not gain system-level privileges.
- Repeated API observations can reveal model behavior or sensitive correlations even without access to the weight files.
Explanation
Correct Answer
D:Independent audit evidence includes: Inspect authentication, rate limits, output granularity, query anomalies and abuse response. The corrective action is: Rate-limit and monitor queries, restrict unnecessary output detail and evaluate extraction risks proportionate to value.
Incorrect Answers
A:This attractive shortcut does not resolve the AI model extraction and excessive query access weakness: Assume model theft requires downloading raw parameters and leave public prediction queries unrestricted. The auditor should instead examine: Inspect authentication, rate limits, output granularity, query anomalies and abuse response.
B:This option concerns AI model rollback and stakeholder communication, which is a related but different AI governance concern. It does not establish control effectiveness for AI model extraction and excessive query access.
C:This option concerns Indirect prompt injection through retrieved content, which is a related but different AI governance concern. It does not establish control effectiveness for AI model extraction and excessive query access.
Question 819
During an AI controls assessment, the team learns that a competitor repeatedly queries a proprietary prediction API to recreate its behavior and pricing decisions. What evidence should the auditor examine FIRST to substantiate or challenge the control owner’s claim? Select ONE best answer.
- Inspect signatures, hash attestations, artifact ACLs, dependency inventory and rebuild records.
- Inspect impact inventory, rollback record, retesting, business communications and lessons-learned owners.
- Assume model theft requires downloading raw parameters and leave public prediction queries unrestricted.
- Inspect authentication, rate limits, output granularity, query anomalies and abuse response.
Explanation
Correct Answer
D:The reason it is necessary is: Repeated API observations can reveal model behavior or sensitive correlations even without access to the weight files. This evidence directly tests the asserted control for AI model extraction and excessive query access: Inspect authentication, rate limits, output granularity, query anomalies and abuse response.
Incorrect Answers
A:This option concerns Foundation model dependency and artifact tampering, which is a related but different AI governance concern. It does not establish control effectiveness for AI model extraction and excessive query access.
B:This option concerns AI model rollback and stakeholder communication, which is a related but different AI governance concern. The decisive principle here is: Repeated API observations can reveal model behavior or sensitive correlations even without access to the weight files.
C:This attractive shortcut does not resolve the AI model extraction and excessive query access weakness: Assume model theft requires downloading raw parameters and leave public prediction queries unrestricted. For this situation: A competitor repeatedly queries a proprietary prediction API to recreate its behavior and pricing decisions.
Question 820
An independent AI auditor reviews this case: A competitor repeatedly queries a proprietary prediction API to recreate its behavior and pricing decisions. Which TWO complementary elements should the auditor seek: an effective corrective control and the most probative independent evidence? Select TWO answers.
- Inspect authentication, rate limits, output granularity, query anomalies and abuse response.
- Rate-limit and monitor queries, restrict unnecessary output detail and evaluate extraction risks proportionate to value.
- Assume model theft requires downloading raw parameters and leave public prediction queries unrestricted.
- Separate development, test and production roles with controlled promotion, scoped credentials and audit trails.
- Inspect tool-call traces, retrieval source trust, instruction hierarchy tests and denied action logs.
Explanation
Correct Answers
A:The reason it is necessary is: Repeated API observations can reveal model behavior or sensitive correlations even without access to the weight files. For this situation: A competitor repeatedly queries a proprietary prediction API to recreate its behavior and pricing decisions.
B:The reason it is necessary is: Repeated API observations can reveal model behavior or sensitive correlations even without access to the weight files. This response fits the risk associated with AI model extraction and excessive query access: Rate-limit and monitor queries, restrict unnecessary output detail and evaluate extraction risks proportionate to value.
Incorrect Answers
C:This attractive shortcut does not resolve the AI model extraction and excessive query access weakness: Assume model theft requires downloading raw parameters and leave public prediction queries unrestricted. Effective closure requires: Suspicious extraction patterns are detected and unauthorized systematic probing is constrained.
D:This option concerns AI data and model CI/CD separation, which is a related but different AI governance concern. It does not establish control effectiveness for AI model extraction and excessive query access.
E:This option concerns Indirect prompt injection through retrieved content, which is a related but different AI governance concern. The decisive principle here is: Repeated API observations can reveal model behavior or sensitive correlations even without access to the weight files.
