Identify Guiding Principles for Responsible AI AI-900 Practice Questions (Page 3)

Questions 11–15

1. 11application · medium · select all that apply

   A bank uses an AI system to decide whether to approve loan applications. The bank wants to be transparent about the system's decisions. Which of the following actions should the bank take? Select all that apply.

   • AAllow customers to request a review of the decision by a human.
   • BKeep the AI system's logic completely secret to protect intellectual property.
   • CProvide customers with the main reasons for a loan decision.
   • DExplain the AI system's general decision-making process in plain language.
   Check answers Discussion Select an answer first
2. 12expert · hard

   A hospital uses an AI system to recommend treatment plans for cancer patients. The system sometimes recommends treatments that are not covered by the patient's insurance. The hospital must balance patient safety, treatment efficacy, and cost. What should the hospital do?

   • AUse the AI's recommendations as final, regardless of insurance coverage, to prioritize efficacy.
   • BHave a human doctor review the AI's recommendations and consider insurance coverage before finalizing the treatment plan.
   • CLet patients decide which treatment to follow after explaining the AI's recommendations.
   • DModify the AI to only recommend treatments that are always covered by insurance.
   Reveal Solution Discussion Select an answer first
4. 13expert · hard

   A global e-commerce company uses an AI chatbot to provide customer support. The chatbot performs well for English-speaking users but often misunderstands users with heavy accents or non-standard dialects. The company wants to make the chatbot more inclusive without alienating its primary English-speaking customer base. What is the best approach?

   • ARequire users to speak slowly and clearly to improve recognition.
   • BUse a more advanced speech recognition model that is known to work well for standard English.
   • CRestrict the chatbot to text-only interactions to avoid accent issues.
   • DTrain the chatbot on a diverse dataset that includes various accents and dialects, and provide a fallback to human agents when the chatbot is uncertain.
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   A city government deploys a chatbot to help residents access public services. The chatbot is text-based and only available in English. The city has a large population of non-English speakers and people with visual impairments. What should the city do to make the chatbot more inclusive?

   • AProvide a phone hotline for non-English speakers and visually impaired users.
   • BUse simpler English to make the chatbot easier for non-native speakers to understand.
   • CKeep the chatbot text-only but add a translation feature for non-English speakers.
   • DAdd support for multiple languages and integrate with screen readers for accessibility.
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   Which responsible AI principle involves designing AI systems to avoid causing harm and to mitigate risks that could affect individuals or society?

   • ATransparency
   • BSafety
   • CPrivacy
   • DAccountability
   Reveal Solution Discussion Select an answer first