AI Concerns, Ethical Principles, and Responsible AI CRAGE Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   Which ethical principle is most directly applied when an organization documents and communicates how an AI system makes decisions?

   • AFairness
   • BTransparency
   • CPrivacy
   • DAccountability
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   Which of the following is a technical concern specifically associated with AI systems?

   • AThe requirement that all AI systems must be connected to the internet to function.
   • BThe tendency of AI systems to consume more electricity than any other computing technology.
   • CThe 'black box' problem, where the decision-making process of complex models is not easily interpretable.
   • DThe inability of AI systems to process large volumes of data.
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A tech company is developing an AI system that will be used by multiple international clients. The company wants to ensure the system respects different cultural norms and legal requirements. Which approach best supports responsible AI in this context?

   • AUse a single global standard for all clients to ensure consistency.
   • BRely on the clients to adapt the system to their local requirements after deployment.
   • CConduct a cross-cultural impact assessment and involve local stakeholders in the design and review process.
   • DDeploy the system only in countries with similar cultural norms to the company's home country.
   Reveal Solution Discussion Select an answer first
5. 4application · medium

   A bank's AI loan approval system is found to deny loans to a higher proportion of applicants from a specific ethnic group. The bank's leadership wants to fix this while maintaining profitability. Which action best balances fairness and business objectives?

   • AIncrease the minimum credit score for all applicants to reduce risk.
   • BRemove all demographic data from the application and retrain the model without any sensitive attributes.
   • CAdjust the model's decision threshold for the affected group to equalize approval rates, then monitor default rates.
   • DLaunch a public relations campaign to explain that the model is fair because it was trained on historical data.
   Reveal Solution Discussion Select an answer first
6. 5application · medium

   A hospital deploys an AI triage system that prioritizes emergency patients. An audit reveals the model assigns lower priority scores to patients from certain zip codes, even after controlling for vital signs. The hospital's ethics committee wants to address this before go-live. Which action best aligns with the responsible AI principle of fairness?

   • AAdd a disclaimer to the model output stating that results may vary by demographic group.
   • BDeploy the model but require a human clinician to override any low-priority assignment for patients over 65.
   • CRetrain the model using a fairness-aware algorithm that enforces equalized odds across protected groups.
   • DRemove zip code from the feature set and deploy the model unchanged.
   Reveal Solution Discussion Select an answer first