Generative AI Leader Fundamentals of Generative AI Practice Question
A financial services firm wants to deploy a generative AI assistant for employees. The assistant must not reveal any customer data and must comply with internal policies. Which Google Cloud capability should they use to control the assistant's responses based on defined safety and privacy rules?
⚠ Common exam trap
It's easy for candidates to confuse infrastructure-level controls like VPC Service Controls or audit logging with content-level guardrails that actually filter model responses.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Vertex AI safety filters and configurable thresholds in the generative AI API.
To enforce safety and privacy rules on generated content, the firm should use Vertex AI's safety filters and configurable thresholds. These settings allow blocking or adjusting responses that violate defined policies, providing a preventive control at the model output level. This directly supports compliance by reducing the chance of leaking customer data or violating internal rules.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Vertex AI safety filters and configurable thresholds in the generative AI API.
Why this is correct
Vertex AI provides safety filters and adjustable thresholds that can block harmful or policy-violating content. By configuring these settings, the firm can reduce the risk of responses that leak sensitive information or violate internal rules. This is the built-in mechanism for applying content safety controls to generative AI responses, making it the appropriate choice for policy enforcement.
- ✗
Cloud Audit Logs to record all API calls made to the assistant.
Why it's wrong here
Cloud Audit Logs provide visibility and forensics by recording administrative and data access events, but they do not prevent or filter responses. They are a detective control, not a preventive one. While valuable for compliance reporting, they cannot stop the assistant from revealing customer data in real time, so they do not satisfy the requirement to control responses.
- ✗
Vertex AI Model Garden to browse and deploy open models.
Why it's wrong here
Model Garden helps discover and deploy models, but it does not enforce custom safety or privacy rules on responses. It is a catalog and deployment surface, not a policy engine. While useful for selecting models, it does not provide the guardrail configuration needed to prevent disclosure of customer data or enforce internal policies, so it does not meet the compliance requirement.
- ✗
VPC Service Controls to create a perimeter around the project.
Why it's wrong here
VPC Service Controls restrict network access to Google Cloud services and help prevent data exfiltration, but they operate at the infrastructure level, not on model outputs. They cannot inspect or filter the content of generated text. The scenario requires controlling what the assistant says, not where it can connect, so this option does not address the response-level policy enforcement.
Go deeper
Related to this question
About these practice questions
One of 1,008 original Generative AI Leader practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Google Cloud exam blueprint
This Generative AI Leader practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the Generative AI Leader exam.