A healthcare startup is developing an AI tool to predict patient readmission risk. The tool will be used by clinicians to allocate follow-up care. The startup has limited resources and must decide where to focus its testing efforts. Which risk assessment approach would best help prioritize testing?

Test the model's accuracy on the most recent patient data available.

Deploy the model in a limited pilot and observe outcomes.

Identify potential failure modes and rate them by likelihood and severity of patient harm.

Review the model's feature importance to understand which inputs drive predictions.

A healthcare AI system monitors patient vitals and alerts staff. The system has been stable for months, but the hospital wants to ensure it continues to perform safely as patient demographics change. The budget is limited. What is the most cost-effective continuous assurance strategy?

Re-train the model monthly with the latest data.

Stop monitoring because the system has been stable.

Conduct a full manual audit every week.

Implement automated monitoring of key performance indicators and data drift, with periodic human review.

A financial services firm is required by a regulator to demonstrate that its AI-based trading system is reliable and compliant. The firm has already performed extensive testing and has documentation of the model's design. What additional activity is most essential to provide ongoing assurance to the regulator?

Establish a continuous monitoring process that tracks the system's performance and compliance indicators over time.

Increase the model's complexity to improve accuracy and reduce risk.

Replace the AI system with a rule-based system to simplify compliance.

Create a static report summarizing the testing results and model design.

An e-commerce company deploys a recommendation model. After three months, the click-through rate drops by 30%. The data science team suspects data drift. What should the assurance team implement to detect this issue early?

Set up continuous monitoring of input data distributions and alert when they deviate from training distributions.

Review the model's training code for bugs.

Re-train the model immediately with the latest data without investigating the cause.

Schedule a one-time audit at the end of the year to review model performance.

EC-CouncilCertified Responsible AI Governance and Ethics

Domain 6Objective 2

AI Assurance, Testing, and Auditing CRAGE Practice Questions (Page 1)

Part of the AI Incident Response, Assurance, and Auditing domain, which makes up ~16% of our current practice bank.

53questions here

11free pages

8concepts

Questions 1–5

1application · medium
A healthcare startup is developing an AI tool to predict patient readmission risk. The tool will be used by clinicians to allocate follow-up care. The startup has limited resources and must decide where to focus its testing efforts. Which risk assessment approach would best help prioritize testing?
Select an answer first
2expert · hard
A healthcare AI system monitors patient vitals and alerts staff. The system has been stable for months, but the hospital wants to ensure it continues to perform safely as patient demographics change. The budget is limited. What is the most cost-effective continuous assurance strategy?
Select an answer first
3application · medium
A financial services firm is required by a regulator to demonstrate that its AI-based trading system is reliable and compliant. The firm has already performed extensive testing and has documentation of the model's design. What additional activity is most essential to provide ongoing assurance to the regulator?
Select an answer first
4application · medium
An e-commerce company deploys a recommendation model. After three months, the click-through rate drops by 30%. The data science team suspects data drift. What should the assurance team implement to detect this issue early?
Select an answer first
5foundation · easy
Which metric is commonly used to evaluate fairness in AI models by measuring the difference in false positive rates between different groups?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to CRAGE

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by EC-Council. “CRAGE” is a trademark of its owner, used for identification only.

AI Assurance, Testing, and Auditing CRAGE Practice Questions (Page 1)

Questions 1–5

A healthcare startup is developing an AI tool to predict patient readmission risk. The tool will be used by clinicians to allocate follow-up care. The startup has limited resources and must decide where to focus its testing efforts. Which risk assessment approach would best help prioritize testing?

A healthcare AI system monitors patient vitals and alerts staff. The system has been stable for months, but the hospital wants to ensure it continues to perform safely as patient demographics change. The budget is limited. What is the most cost-effective continuous assurance strategy?

An e-commerce company deploys a recommendation model. After three months, the click-through rate drops by 30%. The data science team suspects data drift. What should the assurance team implement to detect this issue early?

Which metric is commonly used to evaluate fairness in AI models by measuring the difference in false positive rates between different groups?

Keep studying — AI Incident Response, Assurance, and Auditing