An enterprise machine learning team is implementing automated training pipelines on Google Cloud using managed Vertex AI Custom Training (TrainingPipeline). To protect intellectual property, prevent model weight exfiltration, and enforce the principle of least privilege, the security architect establishes the following requirements: The training job must pull proprietary training container images from Artifact Registry. The training job must stream execution and runtime metrics to Cloud Logging. The training job must write exported model weights and artifacts to a designated Cloud Storage bucket (gs://company-model-registry/output/) without having permission to read, overwrite, or delete existing model files. ML engineers submitting the pipeline must not have direct permissions to read or modify model output files in Cloud Storage, but must be able to launch the job under the approved execution identity. Which identity and access management (IAM) configuration fulfills these requirements?

Assign a dedicated user-managed service account to the training pipeline with `roles/artifactregistry.reader` on the image repository, `roles/logging.logWriter` on the project, and `roles/storage.objectCreator` on the output Cloud Storage bucket. Grant the ML engineers `roles/iam.serviceAccountUser` on the service account and `roles/aiplatform.user` in the project.: ### What the Solution Is This solution establishes a **least-privilege security posture** for fully managed PaaS training workloads (`Vertex AI TrainingPipeline`) by separating the interactive user's identity from a dedicated, fine-grained **user-managed service account** attached to the training job. ### How It Meets the Requirements - **Artifact Registry Access**: Granting the `roles/artifactregistry.reader` role at the specific repository level grants the custom training worker nodes read-only access to pull required container images without allowing modifications or pushes. - **Logging Pipeline Execution**: Granting `roles/logging.logWriter` allows the training environment to output standard runtime logs and execution events directly to **Cloud Logging**. - **Restricting Artifact Export & Preventing Exfiltration**: Granting `roles/storage.objectCreator` on the destination Cloud Storage bucket allows the training script to upload newly generated model weights (via `$AIP_MODEL_DIR`) while strictly denying `storage.objects.get`, `storage.objects.delete`, and list permissions. This prevents the training workload or compromised code from inspecting or exfiltrating existing intellectual property. - **User Impersonation Control**: Granting the ML engineers `roles/iam.serviceAccountUser` (`iam.serviceAccounts.actAs`) on the dedicated service account alongside `roles/aiplatform.user` enables them to launch automated pipelines while completely abstracting direct access to backend storage and repositories. ### Key Advantages and Capabilities - **Separation of Duties**: Isolates developer identity from pipeline actuation permissions. - **Write-Only Bucket Security**: Enforces write-only permissions on sensitive model repositories to prevent unauthorized reads of model weights. - **Targeted IAM Scoping**: Applies permissions at resource-level boundaries (specific bucket and repository) rather than project-wide. - **Immutable Audit Trail**: Ensures all operations and pipeline executions are logged under the dedicated service account identity in Cloud Audit Logs. ### Why This Is Optimal Under Google Cloud's shared responsibility model for managed PaaS training, the platform handles the underlying host VMs, OS hardening, and hypervisors. The customer remains responsible for identity configuration, container security, and data access. Using a dedicated service account with `roles/storage.objectCreator` and `roles/artifactregistry.reader` perfectly enforces zero-trust boundaries without over-granting primitive or administrative roles.

Vertex Ai Pipeline Least Privilege Iam Artifacts | Professional Cloud Security Engineer

Vertex Ai Pipeline Least Privilege Iam Artifacts

Worked example. The correct answer is already marked and every option is explained below, so there is nothing to select here. To answer questions yourself, start the free trial.

An enterprise machine learning team is implementing automated training pipelines on Google Cloud using managed Vertex AI Custom Training (TrainingPipeline). To protect intellectual property, prevent model weight exfiltration, and enforce the principle of least privilege, the security architect establishes the following requirements:

The training job must pull proprietary training container images from Artifact Registry.
The training job must stream execution and runtime metrics to Cloud Logging.
The training job must write exported model weights and artifacts to a designated Cloud Storage bucket (gs://company-model-registry/output/) without having permission to read, overwrite, or delete existing model files.
ML engineers submitting the pipeline must not have direct permissions to read or modify model output files in Cloud Storage, but must be able to launch the job under the approved execution identity.

Which identity and access management (IAM) configuration fulfills these requirements?

Assign the Compute Engine default service account to the training pipeline with roles/editor on the project and roles/storage.objectAdmin on the destination bucket. Grant the ML engineers roles/aiplatform.admin on the project.

Assign the Cloud Build service account to the training pipeline with roles/artifactregistry.repoAdmin and roles/storage.legacyBucketWriter on the bucket. Grant the ML engineers roles/owner on the Vertex AI dataset and model resources.

Assign a dedicated user-managed service account to the training pipeline with roles/artifactregistry.reader on the image repository, roles/logging.logWriter on the project, and roles/storage.objectCreator on the output Cloud Storage bucket. Grant the ML engineers roles/iam.serviceAccountUser on the service account and roles/aiplatform.user in the project.

Assign a dedicated user-managed service account to the training pipeline with roles/artifactregistry.writer, roles/logging.admin, and roles/storage.objectViewer on the destination bucket. Grant the ML engineers roles/storage.admin on the project.

Professional Cloud Security Engineer

Vertex Ai Pipeline Least Privilege Iam Artifacts

Related questions