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 institution is deploying a machine learning model to detect suspicious transactions. Because the institution operates in a strictly regulated industry, leadership has established the following requirements for the model:
Which Google Cloud capability should the institution use to meet these requirements?
Vertex Explainable AI (XAI) is a set of capabilities integrated into Vertex AI and BigQuery ML that helps data scientists, auditors, and business stakeholders understand how machine learning models arrive at their decisions. Instead of treating a complex model as an opaque "black box," Vertex Explainable AI generates feature-based explanations that quantify how much each input factor (such as transaction size, geographic location, or account history) contributed to a specific prediction.
While security controls, logging, and data tracking are essential parts of a holistic AI governance strategy, only Vertex Explainable AI directly inspects the internal decision-making logic of a trained model to explain why a specific output was generated.
Cloud Logging is a centralized, real-time observability service in Google Cloud that collects, stores, and analyzes log data and system events across cloud infrastructure and applications. In the context of AI governance, it supports the Accountability pillar by recording a chronological audit trail of key events, API requests, and final decisions made by AI systems.
The organization needed to establish operational accountability by maintaining a chronological audit trail of system events, user access, and recorded model outputs to monitor overall system performance.
Data lineage is a data governance feature within Knowledge Catalog that automatically records and visualizes the lifecycle of data across an organization. It tracks where datasets originate, which pipelines or systems they pass through, and what transformations are applied to the data along the way.
The organization wanted to audit the journey of its raw data across various cloud storage systems and pipelines to understand what SQL transformations were applied before the data was fed into the model.
Differential privacy is a cryptographic and statistical technique used to protect the privacy of individuals whose records are included in a dataset. When implemented in BigQuery or during machine learning training, it intentionally injects controlled statistical noise into query results or training data so that general patterns can be learned without exposing individual data points.
The organization needed to train a machine learning model or run analytical queries on sensitive customer records while mathematically guaranteeing that no single customer's personal information could be reconstructed or identified from the output.
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