AI Risk Scenario Identification and Assessment (e.g., Threats, Vulnerabilities, and Attacks) AAIR Practice Questions (Page 6)

Questions 26–28

1. 26expert · hard

   An organization has identified four AI risk scenarios with the following likelihood and impact ratings (on a scale of 1-5): Scenario A: likelihood 5, impact 2; Scenario B: likelihood 2, impact 5; Scenario C: likelihood 4, impact 4; Scenario D: likelihood 3, impact 3. The organization has a risk appetite that prioritizes avoiding high-impact risks. Which scenario should be prioritized first?

   • AScenario D
   • BScenario C
   • CScenario B
   • DScenario A
   Reveal Solution Discussion Select an answer first
2. 27application · medium

   A company uses a deep learning model for image recognition in its security system. The model is deployed on edge devices that are physically accessible to the public. An attacker is able to physically tamper with a device and replace the model with a malicious version. Which vulnerability is being exploited?

   • AWeakness in the training process
   • BWeakness in the deployment environment
   • CWeakness in the data pipeline
   • DWeakness in the model architecture
   Reveal Solution Discussion Select an answer first
3. 28foundation · easy

   What is the primary purpose of developing risk scenarios in AI risk management?

   • ATo combine threats, vulnerabilities, and impacts into a realistic narrative for analysis
   • BTo document the AI system's architecture
   • CTo eliminate all risks to the AI system
   • DTo replace the need for quantitative risk analysis
   Reveal Solution Discussion Select an answer first