Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Cloud Digital Leader practice bank has an additional 3,070 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
A financial services organization is building a generative AI application on Google Cloud to assist loan officers in evaluating credit applications. To align with Google's AI Principles and responsible AI practices, the leadership team has established the following ethical requirements:
Which approach should the organization implement to meet these requirements?
Fairness benchmarks and model evaluation metrics are analytical tools used during machine learning development to measure data bias and model bias across protected subgroups—such as gender, race, ethnicity, and religion—helping teams identify potential disparities in model performance.
This concept would be relevant if a team were identifying a known model limitation regarding why an English-evaluated foundation model might require additional multilingual testing before being deployed to a global user base.
To uphold Google's AI Principle to avoid creating or reinforcing unfair bias, organizations must look beyond static pre-deployment benchmarks. True fairness requires evaluating models within the specific demographic and linguistic context of the target audience, paired with continuous post-deployment monitoring.
In Google Cloud's levels of security autonomy, the autonomous level describes a state where AI acts as a trusted system that drives operational lifecycles—such as automated cyber threat detection and response—based on organizational preferences with minimal human intervention.
This approach would be appropriate if the organization were implementing an advanced AI-powered security operations (SecOps) workflow designed to automatically detect and mitigate routine network threats at scale with minimal manual effort.
Under Google's AI Principles, technologies must be accountable to people. This means AI systems must provide appropriate opportunities for feedback, relevant explanations, and appeal, remaining under human direction and control whenever significant personal outcomes are at stake.
Google designs its foundation models and generative AI APIs with Google's AI Principles in mind, incorporating baseline technical safeguards and default content protections. These built-in measures provide a strong starting point for responsible AI development across general use cases.
This option would only be partially applicable if the question asked why Google embeds baseline safety attribute scoring and ethical design into its foundation models prior to customer customization—though even in simple use cases, customer testing and monitoring are still required.
A common misconception is that responsible AI is solely the cloud provider's duty. In reality, responsible AI is a shared responsibility: Google builds models aligned with its AI Principles, while customers must evaluate, test, and govern how those models operate within their specific applications.
Google's AI Principles serve as an ethical framework guiding the development and deployment of artificial intelligence so that it remains socially beneficial, safe, fair, and accountable to people. In practice, operationalizing these principles on Google Cloud requires combining technical safeguards—such as domain-specific tuning, safety filters, and explainability tools—with strong governance and human-in-the-loop oversight.
Unlike approaches that rely solely on pre-configured defaults or complete automation, this holistic strategy combines proactive safety testing, technical guardrails, and human accountability. It directly addresses the inherent limitations of generative AI while upholding ethical standards for high-stakes decision-making.
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