4.3 Describe the Importance of Responsible AI in Business. GENERATIVE-AI-LEADER Practice Questions (Page 5)

Questions 21–25

1. 21application · medium

   A city government is deploying a generative AI chatbot to answer resident questions about public services. Early testing shows the chatbot responds differently to the same question when asked in different dialects, giving less accurate answers for some dialects. What is the most important reason to fix this bias before launch?

   • ATo ensure fair and equitable service for all residents
   • BTo comply with open-source licensing requirements
   • CTo improve the chatbot’s response speed
   • DTo reduce the computational cost of running the chatbot
   Reveal Solution Discussion Select an answer first
2. 22application · medium

   A marketing firm is using a public generative AI API to create personalized ad copy. They plan to paste customer names, email addresses, and purchase histories into the prompts. A privacy analyst warns about data leakage. What is the most effective way to reduce the privacy risk while still using the API?

   • APseudonymize the data by replacing names and emails with random tokens
   • BUse a private, on-premises model that never sends data externally
   • CObtain consent from customers before using their data in the API
   • DAnonymize the data by removing all direct identifiers before sending it to the API
   Reveal Solution Discussion Select an answer first
4. 23foundation · medium

   What is a potential business outcome of using poor-quality data to train an AI system?

   • AIncreased customer satisfaction due to faster model training.
   • BImproved regulatory compliance due to simpler data models.
   • CReduced operational costs due to less data processing.
   • DLoss of customer trust due to incorrect AI-driven decisions.
   Reveal Solution Discussion Select an answer first
5. 24foundation · medium

   What is the key difference between anonymization and pseudonymization?

   • AAnonymization is a type of encryption, while pseudonymization is a type of hashing.
   • BAnonymization is only used for text data, while pseudonymization is used for images.
   • CAnonymization is required by law, while pseudonymization is optional.
   • DAnonymization is irreversible, while pseudonymization can be reversed with additional information.
   Reveal Solution Discussion Select an answer first
6. 25foundation · medium

   Which of the following is a common privacy risk in AI systems?

   • AIncreased model accuracy due to large datasets.
   • BHigher computational cost of training models.
   • CData leakage through model outputs that reveal training data.
   • DReduced need for data storage.
   Reveal Solution Discussion Select an answer first