Generative AI Leader Google Cloud's Generative AI Offerings Practice Question
Which TWO safety features are available in Vertex AI Gemini API? (Select TWO.)
⚠ Common exam trap
A common trap is confusing general Google Cloud security services (like encryption at rest or DLP) with the native safety features of the Vertex AI Gemini API, which only include safety filters and content thresholds.
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
✓
Safety filters for categories like hate speech and harassment
Option A is correct because the Vertex AI Gemini API applies configurable safety filters that screen both prompts and responses against harm categories such as hate speech, harassment, sexually explicit content, and dangerous content. Option B is correct because these safety filters operate using configurable thresholds (for example BLOCK_LOW_AND_ABOVE, BLOCK_MEDIUM_AND_ABOVE, BLOCK_ONLY_HIGH, BLOCK_NONE) that let developers tune how aggressively content is blocked per category. Together, A and B describe the built-in safety attributes exposed through the API's safetySettings. Option C is not a Gemini API safety feature but a general Google Cloud storage/platform control (encryption at rest is handled by the underlying infrastructure, not the API's safety settings). Option D is incorrect because the Gemini API does not automatically redact PII as a built-in safety feature. Option E is incorrect because Cloud DLP is a separate Google Cloud service that must be integrated manually and is not a native safety feature of the Gemini API.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Safety filters for categories like hate speech and harassment
Why this is correct
Safety filters in the Vertex AI Gemini API let you configure thresholds that block or allow content across harm categories including hate speech, harassment, sexually explicit material and dangerous content. This directly satisfies the stem's requirement for available safety features, as the API exposes these configurable filters on requests and responses.
- ✓
Content restrictions based on configurable thresholds
Why this is correct
Configurable thresholds let you set blocking sensitivity per harm category, directly satisfying the API's requirement for tunable safety controls. Vertex AI Gemini's safety settings expose severity levels (for example, block most, block some) that filter both prompts and responses, so content restrictions adjust to your application's tolerance rather than a fixed policy.
- ✗
Model-level encryption at rest
Why it's wrong here
Encryption at rest is a storage-platform control applied to all customer data by default, not a configurable safety feature of the Gemini API's request or response handling. It is tempting because encryption protects data confidentiality, which is the right concern when the requirement is protecting stored training or inference data rather than filtering harmful model input and output.
- ✗
Automatic redaction of personally identifiable information (PII)
Why it's wrong here
The Gemini API's safety features filter harmful content categories such as harassment and dangerous material; they do not automatically detect and redact PII in prompts or responses. It is tempting because PII redaction protects privacy, which is the right control when the requirement is sanitising data before storage or analytics rather than moderating model output.
- ✗
Integration with Cloud Data Loss Prevention (DLP)
Why it's wrong here
Cloud DLP is a separate Sensitive Data Protection service for discovering and de-identifying data in storage and pipelines; the Gemini API does not expose it as a built-in safety setting. It is tempting because DLP addresses sensitive-content risk, which is the correct choice when scanning datasets or data streams for regulated information rather than moderating model prompts and responses.
Go deeper
Related to this question
About these practice questions
This Generative AI Leader question is part of Courseiva's 1,008-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
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.