Question 300
An enterprise audits production agents, but tool calls are logged without the requesting employee's identity. Which design choice best matches the stated constraint?
- Bind actor identity, authorization decision and mutation result in tamper-aware audit records.
- Version prompts and attach deployment identifiers to immutable audit events.
- Combine encryption with least privilege, transport protection, identity controls and data minimization.
- Preserve evidence, decision inputs, model and policy versions and human approval records.
Explanation
Correct Answer
A:A tool result alone cannot establish who initiated an action or whether it was permitted. Map attributable authorization audit evidence to accountable owners, data flows and retained evidence, including connectors and derived AI artifacts.
Incorrect Answers
B:There is a legitimate use case for historical policy reproducibility, but it does not supply the test or enforcement needed for attributable authorization audit evidence. The design needs to bind actor identity, authorization decision and mutation result in tamper-aware audit records.
C:Consider its actual operating scope: multi-layer confidentiality compliance. The requested assurance is attributable authorization audit evidence, which calls for an implementation that can bind actor identity, authorization decision and mutation result in tamper-aware audit records, not merely for a related improvement.
D:The alternative could reduce risk associated with traceable high-impact decision records, yet the described behavior requires attributable authorization audit evidence. That outcome depends on the capability to bind actor identity, authorization decision and mutation result in tamper-aware audit records.
Question 301
A regulator requires rapid deletion of an individual's records, but the team only removes them from the source database. Which architecture adjustment is most appropriate?
- Bind actor identity, authorization decision and mutation result in tamper-aware audit records.
- Apply source-specific lifecycle and deletion controls through the retrieval stores.
- Combine encryption with least privilege, transport protection, identity controls and data minimization.
- Map data flows across models, connectors, caches and logs before selecting approved hosting.
- Implement deletion propagation to derived indexes, caches and applicable retained artifacts.
Explanation
Correct Answer
E:Data subject rights can be defeated by forgotten secondary copies in AI support systems. Map complete deletion propagation to accountable owners, data flows and retained evidence, including connectors and derived AI artifacts.
Incorrect Answers
A:This would make sense as an additional project activity for attributable authorization audit evidence. The priority required here is complete deletion propagation, so the workflow still needs a control that will implement deletion propagation to derived indexes, caches and applicable retained artifacts.
B:By itself, this improves policy-aligned retention in search indexes, but the case is not asking the architect to optimize that dimension. The requirement is complete deletion propagation, supported by an approach to implement deletion propagation to derived indexes, caches and applicable retained artifacts.
C:Its main effect concerns multi-layer confidentiality compliance. The case explicitly hinges on complete deletion propagation; to resolve it reliably, the implementation has to implement deletion propagation to derived indexes, caches and applicable retained artifacts, which this choice does not establish.
D:The option might improve complete residency-aware data flow; that remains insufficient for complete deletion propagation. The case is resolved by complete deletion propagation, which depends on a design that will implement deletion propagation to derived indexes, caches and applicable retained artifacts.
Question 302
A team asserts that encrypting data at rest alone satisfies confidentiality requirements for an AI assistant. Which approach should be recommended?
- Perform a control-by-control gap assessment with legal and security owners.
- Combine encryption with least privilege, transport protection, identity controls and data minimization.
- Implement deletion propagation to derived indexes, caches and applicable retained artifacts.
- Define purpose-limited retention, minimization and access policies for clinical data.
Explanation
Correct Answer
B:Encryption protects certain storage threats but does not stop authorized misuse or overbroad retrieval. Map multi-layer confidentiality compliance to accountable owners, data flows and retained evidence, including connectors and derived AI artifacts.
Incorrect Answers
A:Although documented controls mapped to requirements can be valuable, it is not the acceptance condition in the question. The missing provision is multi-layer confidentiality compliance, addressed by the control to combine encryption with least privilege, transport protection, identity controls and data minimization.
C:It emphasizes complete deletion propagation without closing the stated gap in multi-layer confidentiality compliance. By contrast, the remedy to combine encryption with least privilege, transport protection, identity controls and data minimization acts on the condition that makes this scenario fail.
D:That is primarily a technique for restricted retention of clinical content. It lacks the explicit measure for multi-layer confidentiality compliance that is necessary here, namely a system that will combine encryption with least privilege, transport protection, identity controls and data minimization.
Question 303
A public-sector vendor claims regulatory readiness based only on provider marketing material. What is the best production design response?
- Apply source-specific lifecycle and deletion controls through the retrieval stores.
- Combine encryption with least privilege, transport protection, identity controls and data minimization.
- Map data flows across models, connectors, caches and logs before selecting approved hosting.
- Perform a control-by-control gap assessment with legal and security owners.
Explanation
Correct Answer
D:Vendor compliance statements do not replace validation against a specific organization's actual obligations. Map documented controls mapped to requirements to accountable owners, data flows and retained evidence, including connectors and derived AI artifacts.
Incorrect Answers
A:The alternative could reduce risk associated with policy-aligned retention in search indexes, yet the described behavior requires documented controls mapped to requirements. That outcome depends on the capability to perform a control-by-control gap assessment with legal and security owners.
B:The proposed measure is aimed at multi-layer confidentiality compliance, not documented controls mapped to requirements. The architecture must instead perform a control-by-control gap assessment with legal and security owners; a separate change is still needed for documented controls mapped to requirements.
C:The relevant trade-off is not primarily complete residency-aware data flow. A successful fix needs documented controls mapped to requirements; otherwise documented controls mapped to requirements remains unverified even after this change.
Question 304
A regulated deployment must change prompt policies without losing evidence of which version served each customer. Which change most directly satisfies this requirement?
- Preserve evidence, decision inputs, model and policy versions and human approval records.
- Version prompts and attach deployment identifiers to immutable audit events.
- Bind actor identity, authorization decision and mutation result in tamper-aware audit records.
- Inventory third-party processors and assess transfer mechanisms and contractual duties.
Explanation
Correct Answer
B:Config changes can affect decision outcomes; historical traceability requires preserved policy versions. Map historical policy reproducibility to accountable owners, data flows and retained evidence, including connectors and derived AI artifacts.
Incorrect Answers
A:The option might improve traceable high-impact decision records; that remains insufficient for historical policy reproducibility. The case is resolved by historical policy reproducibility, which depends on a design that will version prompts and attach deployment identifiers to immutable audit events.
C:This addresses attributable authorization audit evidence; however, the problem asks for historical policy reproducibility. Without the safeguard to version prompts and attach deployment identifiers to immutable audit events, the original failure mode is still possible.
D:The proposed measure is aimed at governed third-party data transfers, not historical policy reproducibility. The architecture must instead version prompts and attach deployment identifiers to immutable audit events; a separate change is still needed for historical policy reproducibility.
Question 305
A financial platform is integrating an AI agent but its organization has no agreed incident notification owner. Which solution best resolves this operational gap?
- Define purpose-limited retention, minimization and access policies for clinical data.
- Perform a control-by-control gap assessment with legal and security owners.
- Define response accountability, escalation triggers and evidence preservation before launch.
- Bind actor identity, authorization decision and mutation result in tamper-aware audit records.
Explanation
Correct Answer
C:Compliance is an operational capability as well as a design document; incidents need assigned responsibilities.
Incorrect Answers
A:This addresses restricted retention of clinical content; however, the problem asks for governed incident response obligations. Without the safeguard to define response accountability, escalation triggers and evidence preservation before launch, the original failure mode is still possible.
B:Improving documented controls mapped to requirements does not imply governed incident response obligations. The selected design must define response accountability, escalation triggers and evidence preservation before launch, whereas this choice omits that direct protection or verification step.
D:This is a valid architectural concern when optimizing attributable authorization audit evidence. Here it leaves governed incident response obligations unaddressed, whereas the intended change must define response accountability, escalation triggers and evidence preservation before launch before the case can be considered resolved.
Question 306
An employment assistant appears accurate overall but rejects equivalent applications more often from one population. What design decision would best fit these facts?
- Design accessible interfaces and test assistive-technology workflows.
- Expand representative testing and collect consented feedback from underserved groups.
- Report relevant disaggregated harms, mitigations and limitations to decision makers.
- Measure group-level error disparities and investigate sources before deployment.
Explanation
Correct Answer
D:Aggregate utility can coexist with unequal harm; subgroup testing is necessary to evaluate fairness. Test evidence-based demographic fairness assessment across affected users and give them intelligible disclosures or correction channels where the impact warrants it.
Incorrect Answers
A:The control would be justified if accessible ai user experience were the limiting factor. In this scenario the decisive target is evidence-based demographic fairness assessment; success requires an architecture that will measure group-level error disparities and investigate sources before deployment.
B:The option belongs to the broader work on inclusive language and population coverage, but it cannot by itself deliver evidence-based demographic fairness assessment. A decision on this case needs the team to measure group-level error disparities and investigate sources before deployment and verify that outcome.
C:This shifts attention toward meaningful fairness reporting when the relevant decision is about evidence-based demographic fairness assessment. Unless the architecture can measure group-level error disparities and investigate sources before deployment, the failure described in the stem remains unresolved.
Question 307
A credit-service chatbot recommends decisions without telling users that AI contributed. Which intervention most directly addresses the failure?
- Report relevant disaggregated harms, mitigations and limitations to decision makers.
- Design accessible interfaces and test assistive-technology workflows.
- Provide clear AI involvement disclosures and meaningful ways to request human review.
- Add an appeal and correction channel with accountable escalation.
Explanation
Correct Answer
C:Transparency and user agency support accountability in consequential decision contexts. Test honest disclosure and contestability across affected users and give them intelligible disclosures or correction channels where the impact warrants it.
Incorrect Answers
A:This addresses meaningful fairness reporting; however, the problem asks for honest disclosure and contestability. Without the safeguard to provide clear AI involvement disclosures and meaningful ways to request human review, the original failure mode is still possible.
B:That choice would be responsive to accessible ai user experience, but the reported defect is different. It leaves honest disclosure and contestability unproved because there is no commitment to provide clear AI involvement disclosures and meaningful ways to request human review.
D:Consider its actual operating scope: correctable consequential classifications. The requested assurance is honest disclosure and contestability, which calls for an implementation that can provide clear AI involvement disclosures and meaningful ways to request human review, not merely for a related improvement.
Question 308
A dataset contains mostly metropolitan customer examples, but the system will serve rural dialects too. What should the solution architect prioritize?
- Expand representative testing and collect consented feedback from underserved groups.
- Report relevant disaggregated harms, mitigations and limitations to decision makers.
- Flag uncertainty and provide qualified review for high-impact terms.
- Use transparent data-use notices and lawful, purpose-limited feedback collection.
- Measure group-level error disparities and investigate sources before deployment.
Explanation
Correct Answer
A:Distribution imbalance can produce systematic service gaps that one global benchmark does not reveal. Test inclusive language and population coverage across affected users and give them intelligible disclosures or correction channels where the impact warrants it.
Incorrect Answers
B:This would make sense as an additional project activity for meaningful fairness reporting. The priority required here is inclusive language and population coverage, so the workflow still needs a control that will expand representative testing and collect consented feedback from underserved groups.
C:By itself, this improves transparent limits of machine translation, but the case is not asking the architect to optimize that dimension. The requirement is inclusive language and population coverage, supported by an approach to expand representative testing and collect consented feedback from underserved groups.
D:Its main effect concerns ethical secondary use of interaction data. The case explicitly hinges on inclusive language and population coverage; to resolve it reliably, the implementation has to expand representative testing and collect consented feedback from underserved groups, which this choice does not establish.
E:The option might improve evidence-based demographic fairness assessment; that remains insufficient for inclusive language and population coverage. The case is resolved by inclusive language and population coverage, which depends on a design that will expand representative testing and collect consented feedback from underserved groups.
Question 309
A procurement assistant's explanation uses authoritative language for an uncertain supplier risk score. Which architecture adjustment is most appropriate?
- Present confidence limitations, supporting evidence and disputed assumptions clearly.
- Use transparent data-use notices and lawful, purpose-limited feedback collection.
- Flag uncertainty and provide qualified review for high-impact terms.
- Measure group-level error disparities and investigate sources before deployment.
Explanation
Correct Answer
A:Ethical communication avoids disguising probabilistic judgments as proven facts. Test calibrated transparent risk communication across affected users and give them intelligible disclosures or correction channels where the impact warrants it.
Incorrect Answers
B:Although ethical secondary use of interaction data can be valuable, it is not the acceptance condition in the question. The missing provision is calibrated transparent risk communication, addressed by the control to present confidence limitations, supporting evidence and disputed assumptions clearly.
C:It is aimed at a different outcome, transparent limits of machine translation. No evidence is given that it provides calibrated transparent risk communication; the needed mechanism is to present confidence limitations, supporting evidence and disputed assumptions clearly at the appropriate boundary.
D:It emphasizes evidence-based demographic fairness assessment without closing the stated gap in calibrated transparent risk communication. By contrast, the remedy to present confidence limitations, supporting evidence and disputed assumptions clearly acts on the condition that makes this scenario fail.
Question 310
An agency's AI application is technically accurate but inaccessible to people using screen readers. Which next step is most defensible?
- Design accessible interfaces and test assistive-technology workflows.
- Provide clear AI involvement disclosures and meaningful ways to request human review.
- Use transparent data-use notices and lawful, purpose-limited feedback collection.
- Measure group-level error disparities and investigate sources before deployment.
Explanation
Correct Answer
A:Fair access involves interaction barriers as well as statistical model performance. Test accessible AI user experience across affected users and give them intelligible disclosures or correction channels where the impact warrants it.
Incorrect Answers
B:The alternative could reduce risk associated with honest disclosure and contestability, yet the described behavior requires accessible ai user experience. That outcome depends on the capability to design accessible interfaces and test assistive-technology workflows.
C:The proposed measure is aimed at ethical secondary use of interaction data, not accessible ai user experience. The architecture must instead design accessible interfaces and test assistive-technology workflows; a separate change is still needed for accessible ai user experience.
D:The relevant trade-off is not primarily evidence-based demographic fairness assessment. A successful fix needs accessible ai user experience; otherwise accessible ai user experience remains unverified even after this change.