Palo Alto Networks NetSec-Architect Practice Test 10
Question 90
For an original educational Network Security Architect scenario, consider: A multinational organization enables broad decryption and discovers that certificate errors break important update agents while malware inspection of other flows remains required. Select TWO complementary steps: the targeted correction and its strongest prechange validation.
- Check: Trace the identity mapping, HIP or posture match, device inventory and actual policy rule hit for both a compliant and a noncompliant test endpoint.
- Check: Test certificate chains, decryption policy matches, session failure logs, security-profile coverage and approved exceptions against representative applications.
- Change: Build an explicit inter-tier traffic matrix and enforce least privilege at workload-aware boundaries while accounting for dynamic endpoints and symmetric return paths.
- Force all TLS sessions through decryption without exception and dismiss certificate failures as application defects.
- Change: Design a documented inspection policy with decryption scope, lawful exclusions, certificate trust and compensating controls for unavoidable encrypted blind spots.
Explanation
Correct Answer
B:Security inspection effectiveness depends on content visibility; decryption exemptions need risk-based justification rather than treating all encrypted sessions as equally inspectable. Acceptance: Permitted inspected traffic gains threat coverage while justified exclusions remain functional and are tracked with explicit residual-risk controls.
E:Security inspection effectiveness depends on content visibility; decryption exemptions need risk-based justification rather than treating all encrypted sessions as equally inspectable. Evidence: Test certificate chains, decryption policy matches, session failure logs, security-profile coverage and approved exceptions against representative applications.
Incorrect Answers
A:Check: Trace the identity mapping, HIP or posture match, device inventory and actual policy rule hit for both a compliant and a noncompliant test endpoint. Security inspection effectiveness depends on content visibility; decryption exemptions need risk-based justification rather than treating all encrypted sessions as equally inspectable.
C:Change: Build an explicit inter-tier traffic matrix and enforce least privilege at workload-aware boundaries while accounting for dynamic endpoints and symmetric return paths. A multinational organization enables broad decryption and discovers that certificate errors break important update agents while malware inspection of other flows remains required.
D:Security inspection effectiveness depends on content visibility; decryption exemptions need risk-based justification rather than treating all encrypted sessions as equally inspectable.
Question 91
A network security architect is reviewing this environment: A company intends to deploy a third-party language model inside a customer service assistant, but the risk team has not inspected the model for unsafe behavior or supply-chain exposure. A proposed shortcut is: Deploy an unassessed model and assume ordinary web URL filtering will identify its training-data and behavioral weaknesses. What should the security architect recommend as the narrowly scoped correction? Select ONE best answer.
- Deploy an unassessed model and assume ordinary web URL filtering will identify its training-data and behavioral weaknesses.
- Add a model-assessment stage using appropriate Prisma AIRS AI model scanning and approved remediation or acceptance criteria before production promotion.
- Include controlled AI red-team testing with representative misuse cases, prompt injection challenges, safety benchmarks and documented remediation before enabling privileged integrations.
- Architect AI runtime protection at a supported enforcement point that can inspect prompts and responses for policy-defined risks before sensitive content is released.
Explanation
Correct Answer
B:Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Incorrect Answers
A:Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
C:Include controlled AI red-team testing with representative misuse cases, prompt injection challenges, safety benchmarks and documented remediation before enabling privileged integrations. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
D:Architect AI runtime protection at a supported enforcement point that can inspect prompts and responses for policy-defined risks before sensitive content is released. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Question 92
A Palo Alto Networks architecture review finds: A company intends to deploy a third-party language model inside a customer service assistant, but the risk team has not inspected the model for unsafe behavior or supply-chain exposure. A design team proposes: Add a model-assessment stage using appropriate Prisma AIRS AI model scanning and approved remediation or acceptance criteria before production promotion. Which architecture principle most strongly supports the recommended response? Select ONE best answer.
- Adversarial evaluation probes model and application behavior under malicious prompts, while perimeter firewall tests alone do not expose unsafe agent reasoning or tool permissions.
- Adaptive application identification can improve visibility of changing AI services, but decisions still require an explicit sanctioning and least-privilege policy.
- Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
- An allowed AI application does not imply all data is appropriate to submit; classification and policy must govern each disclosure channel.
Explanation
Correct Answer
C:Relevant validation before approval: Inventory model versions, provenance, scan findings, ownership and deployment gates, then confirm that the scanned artifact matches the deployed release.
Incorrect Answers
A:Adversarial evaluation probes model and application behavior under malicious prompts, while perimeter firewall tests alone do not expose unsafe agent reasoning or tool permissions. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
B:Adaptive application identification can improve visibility of changing AI services, but decisions still require an explicit sanctioning and least-privilege policy. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
D:An allowed AI application does not imply all data is appropriate to submit; classification and policy must govern each disclosure channel. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Question 93
The architecture board has not yet authorized configuration changes. Current evidence indicates: A company intends to deploy a third-party language model inside a customer service assistant, but the risk team has not inspected the model for unsafe behavior or supply-chain exposure. Which evidence best tests the architecture diagnosis before approving a disruptive change? Select ONE best answer.
- Test synthetic sensitive records, matching classifiers, channel visibility, blocked requests, user notification and exception logging.
- Match change tickets to artifact hashes, release configurations, risk assessments, red-team results and actual runtime policy coverage.
- Compare actual application logs, cloud classification, affected tenant URLs, allow and block rules and user experience under a test policy.
- Inventory model versions, provenance, scan findings, ownership and deployment gates, then confirm that the scanned artifact matches the deployed release.
Explanation
Correct Answer
D:Confirm afterward that: Models with unresolved disallowed findings cannot be promoted, while the approved version retains documented scan evidence and exception ownership.
Incorrect Answers
A:Test synthetic sensitive records, matching classifiers, channel visibility, blocked requests, user notification and exception logging. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
B:Match change tickets to artifact hashes, release configurations, risk assessments, red-team results and actual runtime policy coverage. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
C:Compare actual application logs, cloud classification, affected tenant URLs, allow and block rules and user experience under a test policy. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Question 94
The architecture team has implemented an approved correction and must now perform a production-safe acceptance test. Original problem: A company intends to deploy a third-party language model inside a customer service assistant, but the risk team has not inspected the model for unsafe behavior or supply-chain exposure. The operations owner requires positive and negative test evidence and confirmation that legitimate business workflows still function. Which acceptance result demonstrates improved security without breaking authorized workflows? Select ONE best answer.
- The targeted AI function is classified and handled according to business approval while unrelated authorized SaaS workflows continue.
- Approved enterprise AI workflows remain usable and unsanctioned data transfers are logged or prevented per policy.
- Models with unresolved disallowed findings cannot be promoted, while the approved version retains documented scan evidence and exception ownership.
- Protected information is denied or masked according to classification while approved non-sensitive prompts complete normally.
Explanation
Correct Answer
C:Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Incorrect Answers
A:The targeted AI function is classified and handled according to business approval while unrelated authorized SaaS workflows continue. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
B:Approved enterprise AI workflows remain usable and unsanctioned data transfers are logged or prevented per policy. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
D:Protected information is denied or masked according to classification while approved non-sensitive prompts complete normally. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Question 95
A design sign-off request identifies the following problem: A company intends to deploy a third-party language model inside a customer service assistant, but the risk team has not inspected the model for unsafe behavior or supply-chain exposure. Which paired security change and validation should close the design review? Select ONE best answer.
- Change: Discover and classify AI application use, then apply role-aware sanctioned-app controls and documented data sharing restrictions. Verify: Review discovered AI applications, associated users, sanctioned status, DLP event samples and enforcement rules for enterprise versus personal accounts.
- Change: Add a model-assessment stage using appropriate Prisma AIRS AI model scanning and approved remediation or acceptance criteria before production promotion. Verify: Inventory model versions, provenance, scan findings, ownership and deployment gates, then confirm that the scanned artifact matches the deployed release.
- Change: Design an agent boundary with scoped tool permissions, verified user context, untrusted-content isolation and runtime controls for dangerous tool calls. Verify: Test untrusted retrieval content, authorization context, tool invocation logs, deny decisions and the business approval flow for representative high-impact actions.
- Change: Include controlled AI red-team testing with representative misuse cases, prompt injection challenges, safety benchmarks and documented remediation before enabling privileged integrations. Verify: Examine the evaluation plan, attack prompts, guardrail failures, permission scope and retest results for the exact deployed agent release.
Explanation
Correct Answer
B:The change is successful only when: Models with unresolved disallowed findings cannot be promoted, while the approved version retains documented scan evidence and exception ownership.
Incorrect Answers
A:Change: Discover and classify AI application use, then apply role-aware sanctioned-app controls and documented data sharing restrictions.
C:Change: Design an agent boundary with scoped tool permissions, verified user context, untrusted-content isolation and runtime controls for dangerous tool calls.
D:Change: Include controlled AI red-team testing with representative misuse cases, prompt injection challenges, safety benchmarks and documented remediation before enabling privileged integrations.
Question 96
A network security architect is reviewing this environment: A research team imports an externally trained AI model into an internal application and needs a repeatable security gate before new versions enter production. A proposed shortcut is: Deploy an unassessed model and assume ordinary web URL filtering will identify its training-data and behavioral weaknesses. What should the security architect recommend as the narrowly scoped correction? Select ONE best answer.
- Map AI model and application data flows to applicable regulatory obligations and risk controls, then choose architecture and enforcement based on documented requirements.
- Deploy an unassessed model and assume ordinary web URL filtering will identify its training-data and behavioral weaknesses.
- Apply enterprise DLP and AI-aware inspection for sensitive prompt and response content, aligning enforcement with sanctioned applications and documented exceptions.
- Add a model-assessment stage using appropriate Prisma AIRS AI model scanning and approved remediation or acceptance criteria before production promotion.
Explanation
Correct Answer
D:Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Incorrect Answers
A:Map AI model and application data flows to applicable regulatory obligations and risk controls, then choose architecture and enforcement based on documented requirements. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
B:Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
C:Apply enterprise DLP and AI-aware inspection for sensitive prompt and response content, aligning enforcement with sanctioned applications and documented exceptions. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Question 97
A Palo Alto Networks architecture review finds: A research team imports an externally trained AI model into an internal application and needs a repeatable security gate before new versions enter production. A design team proposes: Add a model-assessment stage using appropriate Prisma AIRS AI model scanning and approved remediation or acceptance criteria before production promotion. Which architecture principle most strongly supports the recommended response? Select ONE best answer.
- Availability and risk posture must be planned together; an uninspected backup path or silent fail-open can defeat otherwise correct AI security controls.
- An allowed AI application does not imply all data is appropriate to submit; classification and policy must govern each disclosure channel.
- AI inspection architectures vary between direct application/API integration and network or infrastructure enforcement; a valid design must observe the request/response path it claims to protect.
- Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Explanation
Correct Answer
D:Relevant validation before approval: Inventory model versions, provenance, scan findings, ownership and deployment gates, then confirm that the scanned artifact matches the deployed release.
Incorrect Answers
A:Availability and risk posture must be planned together; an uninspected backup path or silent fail-open can defeat otherwise correct AI security controls. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
B:An allowed AI application does not imply all data is appropriate to submit; classification and policy must govern each disclosure channel. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
C:AI inspection architectures vary between direct application/API integration and network or infrastructure enforcement; a valid design must observe the request/response path it claims to protect. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Question 98
The architecture board has not yet authorized configuration changes. Current evidence indicates: A research team imports an externally trained AI model into an internal application and needs a repeatable security gate before new versions enter production. Which evidence best tests the architecture diagnosis before approving a disruptive change? Select ONE best answer.
- Test untrusted retrieval content, authorization context, tool invocation logs, deny decisions and the business approval flow for representative high-impact actions.
- Compare actual application logs, cloud classification, affected tenant URLs, allow and block rules and user experience under a test policy.
- Inventory model versions, provenance, scan findings, ownership and deployment gates, then confirm that the scanned artifact matches the deployed release.
- Review discovered AI applications, associated users, sanctioned status, DLP event samples and enforcement rules for enterprise versus personal accounts.
Explanation
Correct Answer
C:Confirm afterward that: Models with unresolved disallowed findings cannot be promoted, while the approved version retains documented scan evidence and exception ownership.
Incorrect Answers
A:Test untrusted retrieval content, authorization context, tool invocation logs, deny decisions and the business approval flow for representative high-impact actions. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
B:Compare actual application logs, cloud classification, affected tenant URLs, allow and block rules and user experience under a test policy. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
D:Review discovered AI applications, associated users, sanctioned status, DLP event samples and enforcement rules for enterprise versus personal accounts. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Question 99
The architecture team has implemented an approved correction and must now perform a production-safe acceptance test. Original problem: A research team imports an externally trained AI model into an internal application and needs a repeatable security gate before new versions enter production. The operations owner requires positive and negative test evidence and confirmation that legitimate business workflows still function. Which acceptance result demonstrates improved security without breaking authorized workflows? Select ONE best answer.
- Models with unresolved disallowed findings cannot be promoted, while the approved version retains documented scan evidence and exception ownership.
- Untrusted text cannot cause unauthorized tool execution, and permitted tasks retain auditable identity and change approval.
- Protected information is denied or masked according to classification while approved non-sensitive prompts complete normally.
- The chosen form factor inspects intended AI traffic without creating undisclosed bypass paths or unacceptable application availability risks.
Explanation
Correct Answer
A:Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Incorrect Answers
B:Untrusted text cannot cause unauthorized tool execution, and permitted tasks retain auditable identity and change approval. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
C:Protected information is denied or masked according to classification while approved non-sensitive prompts complete normally. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
D:The chosen form factor inspects intended AI traffic without creating undisclosed bypass paths or unacceptable application availability risks. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
Question 100
For an original educational Network Security Architect scenario, consider: A research team imports an externally trained AI model into an internal application and needs a repeatable security gate before new versions enter production. Select THREE compatible elements: a proper correction, a targeted diagnostic and an acceptance result.
- Result: The design provides evidence of appropriate classification, access, retention and review for applicable rules without assuming a single universal control.
- Change: Add a model-assessment stage using appropriate Prisma AIRS AI model scanning and approved remediation or acceptance criteria before production promotion.
- Result: Models with unresolved disallowed findings cannot be promoted, while the approved version retains documented scan evidence and exception ownership.
- Check: Compare actual application logs, cloud classification, affected tenant URLs, allow and block rules and user experience under a test policy.
- Check: Inventory model versions, provenance, scan findings, ownership and deployment gates, then confirm that the scanned artifact matches the deployed release.
- Change: Apply enterprise DLP and AI-aware inspection for sensitive prompt and response content, aligning enforcement with sanctioned applications and documented exceptions.
Explanation
Correct Answer
B:Evidence: Inventory model versions, provenance, scan findings, ownership and deployment gates, then confirm that the scanned artifact matches the deployed release.
C:Evidence: Inventory model versions, provenance, scan findings, ownership and deployment gates, then confirm that the scanned artifact matches the deployed release. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
E:Acceptance: Models with unresolved disallowed findings cannot be promoted, while the approved version retains documented scan evidence and exception ownership.
Incorrect Answers
A:Result: The design provides evidence of appropriate classification, access, retention and review for applicable rules without assuming a single universal control. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
D:Check: Compare actual application logs, cloud classification, affected tenant URLs, allow and block rules and user experience under a test policy. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.
F:Change: Apply enterprise DLP and AI-aware inspection for sensitive prompt and response content, aligning enforcement with sanctioned applications and documented exceptions. Model scanning assesses risks associated with a model artifact and its behavior before deployment; runtime inspection addresses a different exposure once the model is serving requests.