Palo Alto Networks NetSec-Architect Practice Test 10 - Prepaway
Practice Exams:

Pass Palo Alto Networks NetSec-Architect Exam in First Attempt Guaranteed!

Get 100% Latest Exam Questions, Accurate & Verified Answers!

30 Days Free Updates, Instant Download!

Verified By Experts
Premium File
$69.99 $76.99

NetSec-Architect Premium ETE File


  • Premium File 45 Questions & Answers. Last Update: Nov 10, 2026
Whats Included:
  • Latest Questions
  • 100% Accurate Answers
  • Fast Exam Updates

Palo Alto Networks NetSec-Architect Practice Test 10

Question 90

At fictional Northbridge Network Services, the architecture team is responsible for cross-region ai security design validation during staged modernization and must verify failure behavior and compliance without expanding unnecessary application exposure. Risk case: A multinational AI application has a regional AI runtime policy point but traffic from its secondary failover region bypasses security validation. Verified finding: The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Which scoped architecture decision most directly addresses the documented control gap without unnecessarily changing other services?

  1. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.
  2. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths.
  3. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Architect AI runtime inspection, contextual tool authorization and DLP rules so untrusted retrieved instructions cannot override system-approved capabilities.
  4. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Design an appropriately deployed AI security gateway with supported routed/proxy coverage, identity correlation and policy control for approved AI endpoints.

Explanation

Correct Answer

B:

Because the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls, the relevant step is to extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Security design for AI spans the complete active and failover topology, not only the primary region. Recovery should be checked using this result: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions.

Incorrect Answers

A:

An alternative operation—add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows—does not directly resolve the recorded failure. The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. The appropriate intervention is: Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths.

C:

One alternative is to architect AI runtime inspection, contextual tool authorization and DLP rules so untrusted retrieved instructions cannot override system-approved capabilities, but this does not address the decisive finding: The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Instead, the correction is: Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Check the result using: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions.

D:

An alternative operation—design an appropriately deployed AI security gateway with supported routed/proxy coverage, identity correlation and policy control for approved AI endpoints—does not directly resolve the recorded failure. The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. The appropriate intervention is: Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths.

Question 91

For fictional Northbridge Network Services, architects investigating cross-region ai security design validation during staged modernization observe this risk: A multinational AI application has a regional AI runtime policy point but traffic from its secondary failover region bypasses security validation. Before making broad configuration or topology changes, which finding most directly isolates the failure mechanism?

  1. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset.
  2. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Procurement and processing records lack region, retention and data-flow attributes against the firm's approved governance register.
  3. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls.
  4. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Runtime flow telemetry shows inference pods connecting to a regulated data service outside the application dependency map.

Explanation

Correct Answer

C:

The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. This observation matters because security design for AI spans the complete active and failover topology, not only the primary region. The corresponding targeted response is to extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths.

Incorrect Answers

A:

A different observation—the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset—does not establish the specific fault described here. The record instead shows this: The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Security design for AI spans the complete active and failover topology, not only the primary region.

B:

A competing observation indicates that procurement and processing records lack region, retention and data-flow attributes against the firm's approved governance register, which points away from this case. The distinguishing evidence is: The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. That is consistent with the mechanism: security design for AI spans the complete active and failover topology, not only the primary region.

D:

A different observation—runtime flow telemetry shows inference pods connecting to a regulated data service outside the application dependency map—does not establish the specific fault described here. The record instead shows this: The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Security design for AI spans the complete active and failover topology, not only the primary region.

Question 92

At fictional Northbridge Network Services, the security design for cross-region ai security design validation during staged modernization faces: A multinational AI application has a regional AI runtime policy point but traffic from its secondary failover region bypasses security validation. Diagnostic evidence: The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Proposed design change: Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Which independent observation would demonstrate that the change met its intended outcome?

  1. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, A known noncompliant test model is quarantined, while a signed reviewed model is approved and deployed with traceable evidence.
  2. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions.
  3. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.
  4. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Adversarial test prompts are rejected or contained and authorized customer lookup still works without exporting secrets.

Explanation

Correct Answer

B:

Verify the change by checking: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions. The test is tied to the original clue: the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. The correction being tested is: Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths.

Incorrect Answers

A:

An alternative result would show that A known noncompliant test model is quarantined, while a signed reviewed model is approved and deployed with traceable evidence; it measures a different outcome. To check this fault, verify: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions. The earlier evidence was that the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls.

C:

A different check—benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs—would not demonstrate that this particular failure was corrected. Acceptance instead requires this result: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions. This tests the condition shown by the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls.

D:

An alternative result would show that adversarial test prompts are rejected or contained and authorized customer lookup still works without exporting secrets; it measures a different outcome. To check this fault, verify: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions. The earlier evidence was that the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls.

Question 93

For the architecture board of fictional Northbridge Network Services, a proposed change to cross-region ai security design validation during staged modernization follows this problem: A multinational AI application has a regional AI runtime policy point but traffic from its secondary failover region bypasses security validation. Discriminating observation: The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Which network security architecture principle best explains why the existing design is insufficient?

  1. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, AI governance ties technical controls to jurisdictional data handling, consent and audit obligations.
  2. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, An AI security architecture needs a known enforcement path for every relevant inference request.
  3. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, AI security starts before inference by validating model artifacts and their supply-chain integrity.
  4. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Security design for AI spans the complete active and failover topology, not only the primary region.

Explanation

Correct Answer

D:

Security design for AI spans the complete active and failover topology, not only the primary region. In the reported case, the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. The remedy follows that mechanism: Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths.

Incorrect Answers

A:

The cited mechanism, AI governance ties technical controls to jurisdictional data handling, consent and audit obligations, does not account for the recorded clue. The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Security design for AI spans the complete active and failover topology, not only the primary region.

B:

That alternative rule—an AI security architecture needs a known enforcement path for every relevant inference request—does not explain the deciding observation: the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Security design for AI spans the complete active and failover topology, not only the primary region. That is the mechanism the corrective action must address.

C:

An alternative principle holds that AI security starts before inference by validating model artifacts and their supply-chain integrity, but this is not the mechanism shown here. This case reports that the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Its governing principle is that security design for AI spans the complete active and failover topology, not only the primary region.

Question 94

At fictional Northbridge Network Services, the design team must verify failure behavior and compliance without expanding unnecessary application exposure. Architecture risk: A multinational AI application has a regional AI runtime policy point but traffic from its secondary failover region bypasses security validation. Evidence: The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. Which narrowly scoped design action AND separately measurable postchange confirmation together form the most defensible response?

  1. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Design action: Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Independent proof: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions.
  2. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Design action: Design an appropriately deployed AI security gateway with supported routed/proxy coverage, identity correlation and policy control for approved AI endpoints. Independent proof: Controlled sanctioned model calls are inspected and attributed; unauthorized direct calls are blocked or alerted with evidence.
  3. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Design action: Architect AI runtime inspection, contextual tool authorization and DLP rules so untrusted retrieved instructions cannot override system-approved capabilities. Independent proof: Adversarial test prompts are rejected or contained and authorized customer lookup still works without exporting secrets.
  4. For Northbridge Network Services and the ai inspection path unavailable in failover region assessment, Design action: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Independent proof: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.

Explanation

Correct Answer

A:

The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. The paired correction is to extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Validate its effect with: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions. Security design for AI spans the complete active and failover topology, not only the primary region.

Incorrect Answers

B:

That alternative would test a different outcome: design an appropriately deployed AI security gateway with supported routed/proxy coverage, identity correlation and policy control for approved AI endpoints. Independent proof: Controlled sanctioned model calls are inspected and attributed. Here, the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. The paired response is to extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Its acceptance check is: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions.

C:

At Northbridge Network Services, this pair does not address the recorded clue: the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. The scoped correction is to extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Confirm success using: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions.

D:

The offered action and verification—add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows—do not establish closure of the stated failure. The case evidence is that the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls. The relevant response is to extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Check its effect against: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions.

Question 95

At fictional Cedar Ridge University Systems, the architecture team is responsible for ai security design, continuity, policy and architecture governance and must preserve least privilege, auditable change and measurable business availability. Risk case: A sanctioned AI chat application is permitted for employees, but a test user can upload customer account data into prompts. Verified finding: The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Which scoped architecture decision most directly addresses the documented control gap without unnecessarily changing other services?

  1. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Design an appropriately deployed AI security gateway with supported routed/proxy coverage, identity correlation and policy control for approved AI endpoints.
  2. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths.
  3. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Architect AI runtime inspection, contextual tool authorization and DLP rules so untrusted retrieved instructions cannot override system-approved capabilities.
  4. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

Explanation

Correct Answer

D:

The evidence is that the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The appropriate change is to add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows, because sanctioning an AI app does not eliminate data loss risk from its specific features and transactions. After the change, benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.

Incorrect Answers

A:

One alternative is to design an appropriately deployed AI security gateway with supported routed/proxy coverage, identity correlation and policy control for approved AI endpoints, but this does not address the decisive finding: The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Instead, the correction is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Check the result using: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.

B:

An alternative operation—extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths—does not directly resolve the recorded failure. The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The appropriate intervention is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

C:

The proposed change—architect AI runtime inspection, contextual tool authorization and DLP rules so untrusted retrieved instructions cannot override system-approved capabilities—does not repair the condition documented in the case. The decisive evidence is that the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The necessary change is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

Question 96

For fictional Cedar Ridge University Systems, architects investigating ai security design, continuity, policy and architecture governance observe this risk: A sanctioned AI chat application is permitted for employees, but a test user can upload customer account data into prompts. Before making broad configuration or topology changes, which finding most directly isolates the failure mechanism?

  1. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, The failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls.
  2. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Model registry audit records no signed artifact digest, vulnerability assessment or predeployment model security finding.
  3. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset.
  4. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Test transcripts show a prompt-injection string in lower-trust retrieval content preceding a privileged tool request.

Explanation

Correct Answer

C:

The record shows that the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Sanctioning an AI app does not eliminate data loss risk from its specific features and transactions. The corresponding correction is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

Incorrect Answers

A:

An alternative finding indicates that the failover runbook and route tables show secondary inference endpoints use a direct internet route without equivalent controls; this is not the controlling evidence in the report. The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The appropriate correction is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

B:

A different observation—model registry audit records no signed artifact digest, vulnerability assessment or predeployment model security finding—does not establish the specific fault described here. The record instead shows this: The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Sanctioning an AI app does not eliminate data loss risk from its specific features and transactions.

D:

An alternative finding indicates that test transcripts show a prompt-injection string in lower-trust retrieval content preceding a privileged tool request; this is not the controlling evidence in the report. The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The appropriate correction is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

Question 97

At fictional Cedar Ridge University Systems, the security design for ai security design, continuity, policy and architecture governance faces: A sanctioned AI chat application is permitted for employees, but a test user can upload customer account data into prompts. Diagnostic evidence: The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Proposed design change: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Which independent observation would demonstrate that the change met its intended outcome?

  1. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, A known noncompliant test model is quarantined, while a signed reviewed model is approved and deployed with traceable evidence.
  2. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Controlled sanctioned model calls are inspected and attributed; unauthorized direct calls are blocked or alerted with evidence.
  3. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.
  4. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, A test usage workflow records approved region and retention commitments; disallowed cross-border paths are prevented or escalated.

Explanation

Correct Answer

C:

Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs. This post-change check addresses the recorded finding: the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Sanctioning an AI app does not eliminate data loss risk from its specific features and transactions.

Incorrect Answers

A:

A different check—A known noncompliant test model is quarantined, while a signed reviewed model is approved and deployed with traceable evidence—would not demonstrate that this particular failure was corrected. Acceptance instead requires this result: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs. This tests the condition shown by the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset.

B:

An alternative result would show that controlled sanctioned model calls are inspected and attributed; unauthorized direct calls are blocked or alerted with evidence; it measures a different outcome. To check this fault, verify: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs. The earlier evidence was that the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset.

D:

A different check—A test usage workflow records approved region and retention commitments; disallowed cross-border paths are prevented or escalated—would not demonstrate that this particular failure was corrected. Acceptance instead requires this result: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs. This tests the condition shown by the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset.

Question 98

For the architecture board of fictional Cedar Ridge University Systems, a proposed change to ai security design, continuity, policy and architecture governance follows this problem: A sanctioned AI chat application is permitted for employees, but a test user can upload customer account data into prompts. Discriminating observation: The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Which network security architecture principle best explains why the existing design is insufficient?

  1. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Security design for AI spans the complete active and failover topology, not only the primary region.
  2. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, AI runtime security must cover lateral network flows and workloads, not just prompt text at an API endpoint.
  3. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, An AI security architecture needs a known enforcement path for every relevant inference request.
  4. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Sanctioning an AI app does not eliminate data loss risk from its specific features and transactions.

Explanation

Correct Answer

D:

The governing principle is: Sanctioning an AI app does not eliminate data loss risk from its specific features and transactions. The case specifically shows the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The corresponding change is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

Incorrect Answers

A:

An alternative principle holds that security design for AI spans the complete active and failover topology, not only the primary region, but this is not the mechanism shown here. This case reports that the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Its governing principle is that sanctioning an AI app does not eliminate data loss risk from its specific features and transactions.

B:

The cited mechanism, AI runtime security must cover lateral network flows and workloads, not just prompt text at an API endpoint, does not account for the recorded clue. The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Sanctioning an AI app does not eliminate data loss risk from its specific features and transactions.

C:

That alternative rule—an AI security architecture needs a known enforcement path for every relevant inference request—does not explain the deciding observation: the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Sanctioning an AI app does not eliminate data loss risk from its specific features and transactions. That is the mechanism the corrective action must address.

Question 99

At fictional Cedar Ridge University Systems, the design team must preserve least privilege, auditable change and measurable business availability. Architecture risk: A sanctioned AI chat application is permitted for employees, but a test user can upload customer account data into prompts. Evidence: The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Which narrowly scoped design action AND separately measurable postchange confirmation together form the most defensible response?

  1. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Design action: Design an appropriately deployed AI security gateway with supported routed/proxy coverage, identity correlation and policy control for approved AI endpoints. Independent proof: Controlled sanctioned model calls are inspected and attributed; unauthorized direct calls are blocked or alerted with evidence.
  2. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Design action: Architect AI runtime inspection, contextual tool authorization and DLP rules so untrusted retrieved instructions cannot override system-approved capabilities. Independent proof: Adversarial test prompts are rejected or contained and authorized customer lookup still works without exporting secrets.
  3. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Design action: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Independent proof: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.
  4. For Cedar Ridge University Systems and the sanctioned generative ai leaks regulated records assessment, Design action: Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths. Independent proof: A forced regional switch preserves authorized inference service and blocks the same prohibited test payload in both regions.

Explanation

Correct Answer

C:

The proposed action is to add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Independent verification is: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs. That pair addresses the actual recorded clue: the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset.

Incorrect Answers

A:

This pair does not resolve the recorded clue: the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The scoped correction is to add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Confirm success using: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.

B:

The offered action and verification—architect AI runtime inspection, contextual tool authorization and DLP rules so untrusted retrieved instructions cannot override system-approved capabilities—do not establish closure of the stated failure. The case evidence is that the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The relevant response is to add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Check its effect against: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.

D:

At Cedar Ridge University Systems, this pair does not address the recorded clue: the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The scoped correction is to add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Confirm success using: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.

Question 100

At fictional Atlas Regional Operations, the architecture team is responsible for cross-region ai security design validation during staged modernization and must verify failure behavior and compliance without expanding unnecessary application exposure. Risk case: A sanctioned AI chat application is permitted for employees, but a test user can upload customer account data into prompts. Verified finding: The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Which scoped architecture decision most directly addresses the documented control gap without unnecessarily changing other services?

  1. For Atlas Regional Operations and the sanctioned generative ai leaks regulated records assessment, Extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths.
  2. For Atlas Regional Operations and the sanctioned generative ai leaks regulated records assessment, Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.
  3. For Atlas Regional Operations and the sanctioned generative ai leaks regulated records assessment, Introduce model provenance validation and Prisma AIRS model scanning with approval gates before trusted runtime deployment.
  4. For Atlas Regional Operations and the sanctioned generative ai leaks regulated records assessment, Architect policy tagging, data minimization, regional processing constraints and evidence collection for the selected AI service.

Explanation

Correct Answer

B:

The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. Sanctioning an AI app does not eliminate data loss risk from its specific features and transactions. The direct correction is to add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows. Verify the outcome with: Benign prompts pass while a synthetic regulated identifier upload is blocked and logged without exposing the actual content in logs.

Incorrect Answers

A:

The proposed change—extend approved regional AI security enforcement, routing and policy synchronization to disaster-recovery inference paths—does not repair the condition documented in the case. The decisive evidence is that the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The necessary change is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

C:

An alternative operation—introduce model provenance validation and Prisma AIRS model scanning with approval gates before trusted runtime deployment—does not directly resolve the recorded failure. The approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The appropriate intervention is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

D:

The proposed change—architect policy tagging, data minimization, regional processing constraints and evidence collection for the selected AI service—does not repair the condition documented in the case. The decisive evidence is that the approved SaaS classification exists yet the policy has no context-aware Enterprise DLP classifier for that sensitive dataset. The necessary change is: Add sanctioned-app usage controls and a narrowly scoped Enterprise DLP classifier, with training and authorized exception workflows.

Palo Alto Networks Network Security Architect (NetSec-Architect) Practice Tests