AI Design, Development/Procurement, and Documentation AAIR Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A retail company is procuring a chatbot AI system from a vendor. The vendor's solution depends on a third-party cloud service for natural language processing, and the vendor has a history of frequent service outages. The company requires high availability for customer support. What is the MOST critical procurement risk to address?

   • AThe vendor's user interface design and ease of customization.
   • BThe chatbot's ability to handle multiple languages.
   • CThe vendor's pricing model and total cost of ownership.
   • DThe third-party dependency and its impact on service availability.
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   An AI team is developing a recommendation engine for an e-commerce platform. During the design phase, they document the intended functionality, data sources, and potential risks. What is the PRIMARY purpose of this design documentation?

   • ATo serve as a reference for developers and to support risk assessment and auditability.
   • BTo provide a marketing overview for potential investors.
   • CTo replace the need for testing documentation later in the life cycle.
   • DTo satisfy the vendor's contractual requirements.
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   An AI team is at the problem definition stage of developing a system to predict equipment failures. They have identified that the available sensor data is incomplete for certain machine types. What is the MOST appropriate action at this stage?

   • AImmediately collect more data from all machine types before defining the problem.
   • BRe-evaluate the problem definition and feasibility, considering data availability and potential risks.
   • CReduce the scope of the system to only machine types with complete data.
   • DProceed with model development and address data gaps during deployment.
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   Which AI design principle is most directly concerned with ensuring that the system's decisions can be understood and reviewed by humans?

   • AAccountability
   • BTransparency
   • CRobustness
   • DFairness
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   An insurance company is designing an AI system to automate claim approvals. The design team identifies that the system might inadvertently deny claims more frequently for customers in certain geographic regions due to historical data patterns. What is the MOST effective mitigation strategy at the design phase?

   • AIncorporate fairness constraints into the model design and use diverse training data.
   • BDelay the project until more data is available from all regions.
   • CImplement a post-deployment monitoring system to detect regional disparities.
   • DRemove all geographic features from the model to ensure fairness.
   Reveal Solution Discussion Select an answer first