Google Professional Cloud Security Engineer Practice Test 3
Question 20
A security auditor must review log events without modifying sinks or retention rules. Select TWO statements that correctly combine the Least privilege access to audit logs method with evidence or its governing principle.
- Assign narrowly scoped Cloud Logging view access and separate log configuration permissions.
- Grant project Owner solely to let a contractor search one audit log.
- Correlate supported Cloud Storage Data Access logs with IAM bindings and network context.
- Configure supported Cloud IDS endpoints and Packet Mirroring paths for the intended traffic.
- Test identities can query only their permitted log views and cannot alter audit collection.
Explanation
Correct Answers
A:Assign narrowly scoped Cloud Logging view access and separate log configuration permissions. This is the intended action for Least privilege access to audit logs.
E:Test identities can query only their permitted log views and cannot alter audit collection. This is directly observable validation of the Least privilege access to audit logs result.
Incorrect Answers
B:Grant project Owner solely to let a contractor search one audit log. This is not part of a safe or effective Least privilege access to audit logs design. Giving all developers logging administrator and private-log access may expose sensitive audit metadata.
C:Correlate supported Cloud Storage Data Access logs with IAM bindings and network context. This serves the different Cloud Storage audit trail for exfiltration objective, not the selected Least privilege access to audit logs plan. Log viewer roles and sink or retention administration require distinct privileges.
D:Configure supported Cloud IDS endpoints and Packet Mirroring paths for the intended traffic. This serves the different Cloud Intrusion Detection System alerts objective, not the selected Least privilege access to audit logs plan. Log viewer roles and sink or retention administration require distinct privileges.
Question 21
A chatbot answer sometimes includes sensitive customer identifiers from its retrieved context. Which approach BEST meets the requirement while respecting the system constraints?
- Turn off query logging and assume that alone removes private identifiers from all model outputs.
- Attach a dedicated supported service account and restrict managed training pipeline permissions.
- Apply supported privacy filters, data access segmentation and output inspection for sensitive responses.
- Record trusted artifact provenance, integrity hashes and deployment approvals for model packages.
Explanation
Correct Answer
C:Apply supported privacy filters, data access segmentation and output inspection for sensitive responses. AI outputs can leak memorized or retrieved data unless both inputs and outputs are governed. Verification should demonstrate that test prompts do not expose protected identifiers and redaction events remain auditable.
Incorrect Answers
A:Turn off query logging and assume that alone removes private identifiers from all model outputs. This is an unsafe or ineffective shortcut for AI output data loss prevention. Hiding PII only in the browser DOM does not prevent it from being returned by the API.
B:Attach a dedicated supported service account and restrict managed training pipeline permissions. This primarily addresses Vertex AI training and pipeline service identities, not the specific AI output data loss prevention decision. Here, aI outputs can leak memorized or retrieved data unless both inputs and outputs are governed.
D:Record trusted artifact provenance, integrity hashes and deployment approvals for model packages. This primarily addresses AI model artifact provenance and integrity, not the specific AI output data loss prevention decision. Here, aI outputs can leak memorized or retrieved data unless both inputs and outputs are governed.
Question 22
A triage lead needs consistent ownership and status tracking across project-level findings. An engineer must explain the recommendation for Security Command Center centralized findings. Which underlying principle is the MOST accurate?
- Containment, evidence preservation and recovery are coordinated security operations rather than one dashboard click.
- SCC centralizes supported risk findings but still requires response processes and access configuration.
- Aggregated sinks centralize eligible descendant logs but need correct scope and sink writer identity access.
- Guest OS patching remains an operational responsibility despite cloud-managed physical infrastructure.
Explanation
Correct Answer
B:SCC centralizes supported risk findings but still requires response processes and access configuration. Therefore the implementation decision is to configure Security Command Center at suitable tier and scope with governed finding workflows.
Incorrect Answers
A:Ignoring findings because a dashboard exists leaves exposure unremediated. Containment, evidence preservation and recovery are coordinated security operations rather than one dashboard click.
C:Ignoring findings because a dashboard exists leaves exposure unremediated. Aggregated sinks centralize eligible descendant logs but need correct scope and sink writer identity access.
D:Ignoring findings because a dashboard exists leaves exposure unremediated. This describes the separate VM Manager patch orchestration mechanism; the present case instead requires Security Command Center centralized findings.
Question 23
A payment backend sends personal data through a client connection that requires TLS validation. After the recommended TLS certificate trust and cipher policy change is implemented, which observation MOST directly verifies that it works?
- Output removes intended values and controlled inspection confirms no exposed matches in sample logs.
- Untrusted certificates fail while approved service connections negotiate authenticated encryption.
- Both sides authenticate the expected certificate chain and unauthorized service identities are denied.
- The service respects external key availability and data-access behavior under revocation tests.
Explanation
Correct Answer
B:Untrusted certificates fail while approved service connections negotiate authenticated encryption. This directly tests whether the TLS certificate trust and cipher policy goal was met. Encryption in transit needs endpoint identity validation in addition to encrypted bytes.
Incorrect Answers
A:Output removes intended values and controlled inspection confirms no exposed matches in sample logs. This is evidence about Sensitive Data Protection inspection redaction, not direct proof of the required TLS certificate trust and cipher policy behavior. The relevant check is that untrusted certificates fail while approved service connections negotiate authenticated encryption.
C:Both sides authenticate the expected certificate chain and unauthorized service identities are denied. This is evidence about Mutual TLS for service-to-service identity, not direct proof of the required TLS certificate trust and cipher policy behavior. The relevant check is that untrusted certificates fail while approved service connections negotiate authenticated encryption.
D:The service respects external key availability and data-access behavior under revocation tests. This is evidence about External Key Manager custody, not direct proof of the required TLS certificate trust and cipher policy behavior. The relevant check is that untrusted certificates fail while approved service connections negotiate authenticated encryption.
Question 24
A project retains raw personal records for years after analytics is complete. If the relevant Sensitive data retention and data lifecycle safeguard is omitted, which outcome is the MOST directly associated failure?
- Leaving uncontrolled copies in exports and logs can defeat a primary dataset deletion policy.
- Poor tweak or key management can weaken confidentiality and cause inconsistent decryptability.
- Sharing the underlying dataset viewer role can reveal more than the intended view.
- Blind destruction can trigger permanent data loss in otherwise healthy storage services.
Explanation
Correct Answer
A:Leaving uncontrolled copies in exports and logs can defeat a primary dataset deletion policy. Minimization and lifecycle governance reduce residual sensitive-data exposure beyond IAM encryption. The corrective approach is to test lifecycle execution and backup exception handling against retention requirements.
Incorrect Answers
B:This failure pattern belongs chiefly to Format-preserving encryption data handling; it is not the failure most directly tested in the Sensitive data retention and data lifecycle situation.
C:This failure pattern belongs chiefly to BigQuery authorized views; it is not the failure most directly tested in the Sensitive data retention and data lifecycle situation.
D:This failure pattern belongs chiefly to KMS key version destruction and restore window; it is not the failure most directly tested in the Sensitive data retention and data lifecycle situation.
Question 25
A regulated analytics job must protect data while it is processed in VM memory. If the relevant Confidential VM memory protection safeguard is omitted, which outcome is the MOST directly associated failure?
- Allowing an agent's conversational confidence to authorize high-impact writes bypasses controls.
- Assuming confidential memory automatically blocks malicious application code is unsafe.
- Removing the service agent's key permissions may make resource data unavailable.
- Unavailable external keys can make protected cloud data inaccessible even when IAM allows access.
Explanation
Correct Answer
B:Assuming confidential memory automatically blocks malicious application code is unsafe. Confidential Computing reduces exposure of data in use but does not eliminate the need for identity and encryption controls. The corrective approach is to test supported machine family, image compatibility and confidential mode evidence.
Incorrect Answers
A:Allowing an agent's conversational confidence to authorize high-impact writes bypasses controls. This failure pattern belongs chiefly to Gemini Enterprise Agent Platform tool permission design; it is not the failure most directly tested in the Confidential VM memory protection situation. Assuming confidential memory automatically blocks malicious application code is unsafe.
C:Removing the service agent's key permissions may make resource data unavailable. This failure pattern belongs chiefly to Cloud KMS CMEK service agent access; it is not the failure most directly tested in the Confidential VM memory protection situation. Assuming confidential memory automatically blocks malicious application code is unsafe.
D:Unavailable external keys can make protected cloud data inaccessible even when IAM allows access. This failure pattern belongs chiefly to External Key Manager custody; it is not the failure most directly tested in the Confidential VM memory protection situation. Assuming confidential memory automatically blocks malicious application code is unsafe.
Question 26
A bank wants cryptographic key material governed by an external key management system. After the recommended External Key Manager custody change is implemented, which observation MOST directly verifies that it works?
- Audit evidence attributes reads and writes to intended identity without unrelated dataset access.
- Both sides authenticate the expected certificate chain and unauthorized service identities are denied.
- The service respects external key availability and data-access behavior under revocation tests.
- Untrusted certificates fail while approved service connections negotiate authenticated encryption.
Explanation
Correct Answer
C:The service respects external key availability and data-access behavior under revocation tests. This directly tests whether the External Key Manager custody goal was met. Cloud EKM shifts external key custody and introduces dependency on external key-management availability.
Incorrect Answers
A:Audit evidence attributes reads and writes to intended identity without unrelated dataset access. This is evidence about Vertex AI training and pipeline service identities, not direct proof of the required External Key Manager custody behavior. The relevant check is that the service respects external key availability and data-access behavior under revocation tests.
B:Both sides authenticate the expected certificate chain and unauthorized service identities are denied. This is evidence about Mutual TLS for service-to-service identity, not direct proof of the required External Key Manager custody behavior. The relevant check is that the service respects external key availability and data-access behavior under revocation tests.
D:Untrusted certificates fail while approved service connections negotiate authenticated encryption. This is evidence about TLS certificate trust and cipher policy, not direct proof of the required External Key Manager custody behavior. The relevant check is that the service respects external key availability and data-access behavior under revocation tests.
Question 27
One VPN tunnel is down but regulated workloads require continued private connectivity. Select THREE statements that correctly combine the HA VPN encrypted hybrid tunnels method with evidence or its governing principle.
- Use a single static public TCP proxy and call it a redundant encrypted site-to-site VPN.
- Traffic remains encrypted and routes converge when a single tunnel or interface fails.
- HA VPN provides IPsec-based encrypted connectivity while redundancy depends on correct tunnel and BGP design.
- Configure supported restricted.googleapis.com DNS and routing where the perimeter design requires it.
- Configure HA VPN with redundant interfaces and Cloud Router BGP sessions across supported gateways.
- Configure Cloud NAT with appropriate gateway, addresses and subnet egress coverage.
Explanation
Correct Answers
B:Traffic remains encrypted and routes converge when a single tunnel or interface fails. This is directly observable validation of the HA VPN encrypted hybrid tunnels result.
C:HA VPN provides IPsec-based encrypted connectivity while redundancy depends on correct tunnel and BGP design. This explains why the implementation and evidence fit together.
E:Configure HA VPN with redundant interfaces and Cloud Router BGP sessions across supported gateways. This is the intended action for HA VPN encrypted hybrid tunnels.
Incorrect Answers
A:Use a single static public TCP proxy and call it a redundant encrypted site-to-site VPN. This is not part of a safe or effective HA VPN encrypted hybrid tunnels design. One misconfigured BGP peer can eliminate the claimed high-availability path.
D:Configure supported restricted.googleapis.com DNS and routing where the perimeter design requires it. This serves the different Restricted Google APIs VIP for perimeters objective, not the selected HA VPN encrypted hybrid tunnels plan. HA VPN provides IPsec-based encrypted connectivity while redundancy depends on correct tunnel and BGP design.
F:Configure Cloud NAT with appropriate gateway, addresses and subnet egress coverage. This serves the different Cloud NAT outbound without inbound exposure objective, not the selected HA VPN encrypted hybrid tunnels plan. HA VPN provides IPsec-based encrypted connectivity while redundancy depends on correct tunnel and BGP design.
Question 28
A service account can read a resource in project A but fails in project B despite similar role names. If the relevant IAM effective permission troubleshooting safeguard is omitted, which outcome is the MOST directly associated failure?
- Orphan accounts and retained group memberships can expose projects after employment ends.
- Guessing from only the project IAM screen may overlook higher-level policies.
- Broad primitive roles can expose multiple projects and protected datasets unnecessarily.
- A malformed condition or unsupported attribute may broaden or unexpectedly block access.
Explanation
Correct Answer
B:Guessing from only the project IAM screen may overlook higher-level policies. Effective IAM authorization depends on inheritance, permission, conditions and explicit deny interactions. The corrective approach is to reproduce the authorization check with precise principal, permission and target resource.
Incorrect Answers
A:Orphan accounts and retained group memberships can expose projects after employment ends. This failure pattern belongs chiefly to Cloud Identity user lifecycle automation; it is not the failure most directly tested in the IAM effective permission troubleshooting situation. Guessing from only the project IAM screen may overlook higher-level policies.
C:Broad primitive roles can expose multiple projects and protected datasets unnecessarily. This failure pattern belongs chiefly to IAM least-privilege role scoping; it is not the failure most directly tested in the IAM effective permission troubleshooting situation. Guessing from only the project IAM screen may overlook higher-level policies.
D:A malformed condition or unsupported attribute may broaden or unexpectedly block access. This failure pattern belongs chiefly to Conditional IAM access restrictions; it is not the failure most directly tested in the IAM effective permission troubleshooting situation. Guessing from only the project IAM screen may overlook higher-level policies.
Question 29
Security finds an unused Google Cloud API enabled in a high-sensitivity project. If the relevant API exposure restriction and service inventory safeguard is omitted, which outcome is the MOST directly associated failure?
- Treating an allow-on-443 rule as deep content validation leaves application threats unexamined.
- Incorrect DS rollover can cause validating resolvers to reject legitimate zone responses.
- Using the private VIP indiscriminately can allow calls to services outside the intended restricted set.
- Disabling an API used by dependent managed services can disrupt workloads.
Explanation
Correct Answer
D:Disabling an API used by dependent managed services can disrupt workloads. Enabled service APIs and permission to call them are distinct controls. The corrective approach is to check actual API use and service dependencies before introducing restrictions.
Incorrect Answers
A:Treating an allow-on-443 rule as deep content validation leaves application threats unexamined. This failure pattern belongs chiefly to Cloud NGFW application-layer inspection; it is not the failure most directly tested in the API exposure restriction and service inventory situation. Disabling an API used by dependent managed services can disrupt workloads.
B:Incorrect DS rollover can cause validating resolvers to reject legitimate zone responses. This failure pattern belongs chiefly to Cloud DNSSEC authoritative zone protection; it is not the failure most directly tested in the API exposure restriction and service inventory situation. Disabling an API used by dependent managed services can disrupt workloads.
C:This failure pattern belongs chiefly to Restricted Google APIs VIP for perimeters; it is not the failure most directly tested in the API exposure restriction and service inventory situation.
Question 30
Security needs a pipeline rule preventing unsafe cloud configurations before apply. Which practical correction or implementation decision is MOST directly appropriate?
- Rebuild and modify the test image to confirm signature enforcement and approver separation.
- Test known-bad infrastructure fixtures and validate exemptions or approvals.
- Test pod identities from distinct namespaces and confirm restricted API calls fail.
- Trust every Terraform apply simply because its syntax is valid.
Explanation
Correct Answer
B:Test known-bad infrastructure fixtures and validate exemptions or approvals. The intended behavior can be checked because an intentionally unsafe plan fails the gate while a least-privilege revision passes.
Incorrect Answers
A:Rebuild and modify the test image to confirm signature enforcement and approver separation. This could help Binary Authorization attestor chain of custody, but it does not resolve the Infrastructure as code security validation case. IaC policy validation can catch configuration drift earlier than production vulnerability alerts.
C:Test pod identities from distinct namespaces and confirm restricted API calls fail. This could help GKE Workload Identity for deployment runtime, but it does not resolve the Infrastructure as code security validation case. IaC policy validation can catch configuration drift earlier than production vulnerability alerts.
D:Trust every Terraform apply simply because its syntax is valid. This would not correct the Infrastructure as code security validation workflow. A review comment without enforced CI policy can still permit unsafe infrastructure rollout.