Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Professional Cloud Architect practice bank has 2,335 more like it. Try 10 free: instant score, no account.
A financial enterprise is deploying a customer-facing conversational application powered by foundation large language models (LLMs) on Google Cloud. During security assessments, red-team testers successfully executed sophisticated prompt injection and jailbreaking attacks that bypassed standard system prompts and caused the model to exhibit unintended behaviors.
The enterprise has the following operational requirements:
Which security capability should the cloud architect implement to meet these requirements?
Default non-configurable safety filters provide an automated baseline layer of safety built directly into foundation models like Gemini. Their primary function is to enforce essential legal and ethical baselines without requiring explicit configuration.
The workload were an internal, low-risk demonstration environment where no malicious or untrusted user input is expected and baseline CSAM/copyright protection is sufficient.
System instructions provide direct behavioral guidance and persona framing to a model before processing incoming user prompts. They define preferred communication styles, task constraints, tone, and operational boundaries.
The goal was simply to establish brand tone, conversational context, or general behavioral guidelines for normal user interactions without defending against active security exploits.
Google Cloud Armor is a network edge and application perimeter security service that integrates with Google Cloud External Application Load Balancers. It provides Layer 3 to Layer 7 protection against Distributed Denial of Service (DDoS) attacks and common web exploits (such as SQL injection and Cross-Site Scripting via OWASP Top 10 rules).
The primary requirement were protecting the web frontend and API gateways from network DDoS attacks, IP-based scraping, or traditional SQL/XSS injections.
Model Armor is an enterprise-grade security service designed specifically for generative AI workloads and foundation models in Google Cloud. It acts as an operational guardrail layer that evaluates both user inputs (prompts) and generated outputs to safeguard LLM interactions against specialized AI threats.

Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated