Google Professional Cloud Security Engineer Practice Test 72
Question 710
An automation service maintains a security group for a rotating set of approved analysts. If the relevant Programmatic group membership governance safeguard is omitted, which outcome is the MOST directly associated failure?
- Allowing any deployer to attach an owner-level service account can enable privilege escalation.
- Unverified group automation may silently grant inherited IAM roles to unintended principals.
- Routine use of shared super-admin credentials destroys accountability and enlarges blast radius.
- Checking only successful SSO sessions overlooks excess permissions inherited through groups.
Explanation
Correct Answer
B:Unverified group automation may silently grant inherited IAM roles to unintended principals. Managing authorization through groups can simplify changes but group administration itself is privileged. The corrective approach is to reconcile approved roster changes against group membership and resulting project access.
Incorrect Answers
A:This failure pattern belongs chiefly to Scoped permission for service account creation; it is not the failure most directly tested in the Programmatic group membership governance situation.
C:Routine use of shared super-admin credentials destroys accountability and enlarges blast radius. This failure pattern belongs chiefly to Super administrator emergency access; it is not the failure most directly tested in the Programmatic group membership governance situation. Unverified group automation may silently grant inherited IAM roles to unintended principals.
D:Checking only successful SSO sessions overlooks excess permissions inherited through groups. This failure pattern belongs chiefly to Group-based provisioning audit trail; it is not the failure most directly tested in the Programmatic group membership governance situation. Unverified group automation may silently grant inherited IAM roles to unintended principals.
Question 711
A machine learning team imports a pretrained model artifact from a third-party repository. Which approach BEST meets the requirement while respecting the system constraints?
- Attach a dedicated supported service account and restrict managed training pipeline permissions.
- Trust any public model artifact because its README contains the word verified.
- Configure least-privilege agent identity and connector permissions with explicit action approval gates.
- Record trusted artifact provenance, integrity hashes and deployment approvals for model packages.
Explanation
Correct Answer
D:Record trusted artifact provenance, integrity hashes and deployment approvals for model packages. Model weights and packages are supply-chain artifacts subject to integrity and provenance controls. Verification should demonstrate that deployed artifacts match reviewed digests and unauthorized version swaps are detected.
Incorrect Answers
A: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 model artifact provenance and integrity decision. Here, model weights and packages are supply-chain artifacts subject to integrity and provenance controls.
B:Trust any public model artifact because its README contains the word verified. This is an unsafe or ineffective shortcut for AI model artifact provenance and integrity. Importing unreviewed model files can introduce malicious code or poisoned behavior.
C:Configure least-privilege agent identity and connector permissions with explicit action approval gates. This primarily addresses Gemini Enterprise Agent Platform tool permission design, not the specific AI model artifact provenance and integrity decision. Here, model weights and packages are supply-chain artifacts subject to integrity and provenance controls.
Question 712
A Cloud SQL application needs database access differentiated from the Google Cloud project's IAM editor role. An engineer must explain the recommendation for Cloud SQL private IP and database authentication. Which underlying principle is the MOST accurate?
- Network connectivity and database-level authentication are separate enforcement layers.
- Format-preserving encryption retains structural constraints but is not the same as plain-text masking.
- Metadata server tokens reflect attached workload identity and must be protected from untrusted code paths.
- IaaS-hosted models leave more guest and runtime controls with customers than managed AI services.
Explanation
Correct Answer
A:Network connectivity and database-level authentication are separate enforcement layers. Therefore the implementation decision is to configure Cloud SQL private connectivity and a supported database authentication model with least privilege.
Incorrect Answers
B:A project IAM grant alone does not necessarily authorize arbitrary SQL queries on every database. Format-preserving encryption retains structural constraints but is not the same as plain-text masking.
C:A project IAM grant alone does not necessarily authorize arbitrary SQL queries on every database. Metadata server tokens reflect attached workload identity and must be protected from untrusted code paths.
D:A project IAM grant alone does not necessarily authorize arbitrary SQL queries on every database. This describes the separate IaaS versus managed AI training boundary mechanism; the present case instead requires Cloud SQL private IP and database authentication.
Question 713
A security engineer needs to deactivate one suspicious service account during investigation. Which practical correction or implementation decision is MOST directly appropriate?
- Validate key disablement through audit logs and test legitimate replacement authentication.
- Change the service account display name and assume all tokens immediately become invalid.
- Review runtime identity bindings and test legitimate and prohibited API calls.
- Inspect recent authentication activity and test the replacement runtime principal before retirement.
Explanation
Correct Answer
D:Inspect recent authentication activity and test the replacement runtime principal before retirement. The intended behavior can be checked because unauthorized API calls stop and explicitly approved dependent workloads are migrated or remediated.
Incorrect Answers
A:Validate key disablement through audit logs and test legitimate replacement authentication. This could help Service account key inventory and exposure response, but it does not resolve the Service account disabling and dependency mapping case. Disabling an identity affects all callers using it, not just one label or attached role.
B:Change the service account display name and assume all tokens immediately become invalid. This would not correct the Service account disabling and dependency mapping workflow. Blindly deleting a shared service account can cause unplanned production outages.
C:Review runtime identity bindings and test legitimate and prohibited API calls. This could help Service account per-workload identity design, but it does not resolve the Service account disabling and dependency mapping case. Disabling an identity affects all callers using it, not just one label or attached role.
Question 714
Security needs to verify that an approved model version is the one deployed to production. Which practical correction or implementation decision is MOST directly appropriate?
- Create a workload-specific shared-responsibility matrix and verify controls.
- Validate source authenticity, digest and permitted registry access during release.
- Test tool permission matrix, approval workflow and invocation audit trail.
- Trust any public model artifact because its README contains the word verified.
Explanation
Correct Answer
B:Validate source authenticity, digest and permitted registry access during release. The intended behavior can be checked because deployed artifacts match reviewed digests and unauthorized version swaps are detected.
Incorrect Answers
A:Create a workload-specific shared-responsibility matrix and verify controls. This could help IaaS versus managed AI training boundary, but it does not resolve the AI model artifact provenance and integrity case. Model weights and packages are supply-chain artifacts subject to integrity and provenance controls.
C:Test tool permission matrix, approval workflow and invocation audit trail. This could help Gemini Enterprise Agent Platform tool permission design, but it does not resolve the AI model artifact provenance and integrity case. Model weights and packages are supply-chain artifacts subject to integrity and provenance controls.
D:Trust any public model artifact because its README contains the word verified. This would not correct the AI model artifact provenance and integrity workflow. Importing unreviewed model files can introduce malicious code or poisoned behavior.
Question 715
A business wants temporary exports removed automatically but litigation evidence must be retained. If the relevant Legal hold versus lifecycle delete safeguard is omitted, which outcome is the MOST directly associated failure?
- A broad expiration rule without holds can conflict with retention obligations.
- An overly permissive domain exception can enable unauthorized external role assignments.
- Calling default encryption customer-owned without actual customer key management is misleading.
- Omitting customer-owned guest patches or data authorization creates compliance gaps.
Explanation
Correct Answer
A:A broad expiration rule without holds can conflict with retention obligations. Legal preservation and ordinary object lifecycle deletion are separate data governance requirements. The corrective approach is to test retention, hold release approval and lifecycle behavior on nonproduction objects.
Incorrect Answers
B:An overly permissive domain exception can enable unauthorized external role assignments. This failure pattern belongs chiefly to Restricting allowed IAM member domains; it is not the failure most directly tested in the Legal hold versus lifecycle delete situation. A broad expiration rule without holds can conflict with retention obligations.
C:Calling default encryption customer-owned without actual customer key management is misleading. This failure pattern belongs chiefly to Encryption compliance and external key governance; it is not the failure most directly tested in the Legal hold versus lifecycle delete situation. A broad expiration rule without holds can conflict with retention obligations.
D:Omitting customer-owned guest patches or data authorization creates compliance gaps. This failure pattern belongs chiefly to Cloud shared responsibility service boundaries; it is not the failure most directly tested in the Legal hold versus lifecycle delete situation. A broad expiration rule without holds can conflict with retention obligations.
Question 716
A compliance team plans to destroy the last usable version of a key protecting retained records. Which practical correction or implementation decision is MOST directly appropriate?
- Treat a destroyed encryption key version as a password reset that always restores old data.
- Inventory encrypted resource dependencies and test decrypt before retiring old versions.
- Test match conditions and retention interaction on sample objects before enforcement.
- Test key disable and recovery scenario on nonproduction encrypted data before scheduling destruction.
Explanation
Correct Answer
D:Test key disable and recovery scenario on nonproduction encrypted data before scheduling destruction. The intended behavior can be checked because data access behavior and any permitted restore window are understood before irreversible destruction.
Incorrect Answers
A:Treat a destroyed encryption key version as a password reset that always restores old data. This would not correct the KMS key version destruction and restore window workflow. Blind destruction can trigger permanent data loss in otherwise healthy storage services.
B:Inventory encrypted resource dependencies and test decrypt before retiring old versions. This could help Customer-managed key rotation policy, but it does not resolve the KMS key version destruction and restore window case. Cryptographic key destruction can make encrypted data unrecoverable regardless of object retention.
C:Test match conditions and retention interaction on sample objects before enforcement. This could help Cloud Storage object lifecycle transition, but it does not resolve the KMS key version destruction and restore window case. Cryptographic key destruction can make encrypted data unrecoverable regardless of object retention.
Question 717
A business wants to prevent employees from granting project roles to arbitrary outside identities. If the relevant Restricting allowed IAM member domains safeguard is omitted, which outcome is the MOST directly associated failure?
- An overly permissive domain exception can enable unauthorized external role assignments.
- Broad unreviewed deletion rules can disable legitimate cloud users during synchronization.
- Misplaced high-level grants can grant permissions to every descendant project.
- Putting development projects in a production folder can impose unintended controls and privileges.
Explanation
Correct Answer
A:An overly permissive domain exception can enable unauthorized external role assignments. Domain-restricted sharing guardrails constrain identity grants separately from resource firewalls. The corrective approach is to test allowed and blocked external principals and record exception governance.
Incorrect Answers
B:Broad unreviewed deletion rules can disable legitimate cloud users during synchronization. This failure pattern belongs chiefly to Cloud Identity directory synchronization; it is not the failure most directly tested in the Restricting allowed IAM member domains situation. An overly permissive domain exception can enable unauthorized external role assignments.
C:Misplaced high-level grants can grant permissions to every descendant project. This failure pattern belongs chiefly to Organization hierarchy and inherited roles; it is not the failure most directly tested in the Restricting allowed IAM member domains situation. An overly permissive domain exception can enable unauthorized external role assignments.
D:Putting development projects in a production folder can impose unintended controls and privileges. This failure pattern belongs chiefly to Project folder segmentation for environments; it is not the failure most directly tested in the Restricting allowed IAM member domains situation. An overly permissive domain exception can enable unauthorized external role assignments.
Question 718
Security wants changes in a sandbox to inherit different policies from production. Select TWO statements that correctly combine the Project folder segmentation for environments method with evidence or its governing principle.
- Tests confirm production policies apply only to intended descendants and sandbox exceptions remain scoped.
- Define and enforce an applicable custom organization policy constraint with supported resource rules.
- Change a project display name to Production and assume it moves into the production folder.
- Use folders and project placement aligned with environment boundaries and inherited guardrails.
- Grant only the narrow permitted policy exception and retain documented rollback and audit approval.
Explanation
Correct Answers
A:Tests confirm production policies apply only to intended descendants and sandbox exceptions remain scoped. This is directly observable validation of the Project folder segmentation for environments result.
D:Use folders and project placement aligned with environment boundaries and inherited guardrails. This is the intended action for Project folder segmentation for environments.
Incorrect Answers
B:Define and enforce an applicable custom organization policy constraint with supported resource rules. This serves the different Custom organization policy on resource attributes objective, not the selected Project folder segmentation for environments plan. Resource hierarchy shapes both inheritance and administrative blast radius.
C:Change a project display name to Production and assume it moves into the production folder. This is not part of a safe or effective Project folder segmentation for environments design. Putting development projects in a production folder can impose unintended controls and privileges.
E:Grant only the narrow permitted policy exception and retain documented rollback and audit approval. This serves the different Hierarchy-scoped exception and policy rollback objective, not the selected Project folder segmentation for environments plan. Resource hierarchy shapes both inheritance and administrative blast radius.
Question 719
Security wants a distinct team to control who may attach high-privilege service accounts to new resources. Select TWO statements that correctly combine the Scoped permission for service account creation method with evidence or its governing principle.
- Create distinct runtime service accounts and bind task-specific roles at the smallest supported scope.
- Grant project Owner to every CI developer solely to let them select a service account.
- Inventory workloads using the identity and disable or suspend the affected service account with a recovery plan.
- Separate service account creation, service account user and IAM policy administration privileges.
- Developers can attach only designated execution identities and cannot elevate by choosing unrelated accounts.
Explanation
Correct Answers
D:Separate service account creation, service account user and IAM policy administration privileges. This is the intended action for Scoped permission for service account creation.
E:Developers can attach only designated execution identities and cannot elevate by choosing unrelated accounts. This is directly observable validation of the Scoped permission for service account creation result.
Incorrect Answers
A:Create distinct runtime service accounts and bind task-specific roles at the smallest supported scope. This serves the different Service account per-workload identity design objective, not the selected Scoped permission for service account creation plan. The permission to act as a service account is distinct from permission to edit the workload configuration.
B:Grant project Owner to every CI developer solely to let them select a service account. This is not part of a safe or effective Scoped permission for service account creation design. Allowing any deployer to attach an owner-level service account can enable privilege escalation.
C:This serves the different Service account disabling and dependency mapping objective, not the selected Scoped permission for service account creation plan.
Question 720
A VM boots a tampered bootloader after an unauthorized base image modification. If the relevant Shielded VM secure boot integrity safeguard is omitted, which outcome is the MOST directly associated failure?
- Ignoring scan results until after deployment reduces time to remediate vulnerable releases.
- Giving all developers logging administrator and private-log access may expose sensitive audit metadata.
- Disabling integrity monitoring can hide unauthorized boot image changes.
- Untested all-fleet patching may cause application regressions or leave unpatched instances unnoticed.
Explanation
Correct Answer
C:Disabling integrity monitoring can hide unauthorized boot image changes. Shielded VM features protect portions of the boot process but do not replace guest patching or IAM. The corrective approach is to test secure boot compatibility and review VM integrity monitoring events.
Incorrect Answers
A:Ignoring scan results until after deployment reduces time to remediate vulnerable releases. This failure pattern belongs chiefly to Continuous integration dependency security gate; it is not the failure most directly tested in the Shielded VM secure boot integrity situation. Disabling integrity monitoring can hide unauthorized boot image changes.
B:Giving all developers logging administrator and private-log access may expose sensitive audit metadata. This failure pattern belongs chiefly to Least privilege access to audit logs; it is not the failure most directly tested in the Shielded VM secure boot integrity situation. Disabling integrity monitoring can hide unauthorized boot image changes.
D:Untested all-fleet patching may cause application regressions or leave unpatched instances unnoticed. This failure pattern belongs chiefly to VM Manager patch orchestration; it is not the failure most directly tested in the Shielded VM secure boot integrity situation. Disabling integrity monitoring can hide unauthorized boot image changes.