Security Controls and Monitoring AAISM Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   Which method is commonly used to continuously monitor an AI model for data drift?

   • AReviewing the model's source code for vulnerabilities
   • BChecking the model's accuracy on historical test data
   • CMonitoring the model's response time
   • DComparing the distribution of incoming data to the training data distribution
   Reveal Solution Discussion Select an answer first
2. 22application · medium

   A company's AI-based image recognition system suddenly misclassifies stop signs as yield signs in a production environment. The security team suspects an adversarial attack. What immediate action should the team take to confirm the suspicion?

   • ACheck the model's training data for signs of data poisoning.
   • BIncrease the model's confidence threshold for all classifications.
   • CRebuild the model from scratch using a different training dataset.
   • DAnalyze the input images for subtle perturbations that are imperceptible to humans.
   Reveal Solution Discussion Select an answer first
4. 23expert · hard

   A company's AI-powered customer service chatbot has been manipulated to reveal sensitive customer information. The incident response team is activated. The team must decide on a response that minimizes customer impact while preserving evidence for legal action. What is the best course of action?

   • ANotify all customers immediately before any investigation.
   • BContinue operating the chatbot while investigating to avoid customer disruption.
   • CImmediately shut down the chatbot and delete all logs to prevent further exposure.
   • DIsolate the chatbot, preserve logs, and notify affected customers after a preliminary investigation.
   Reveal Solution Discussion Select an answer first
5. 24application · medium

   An organization is preparing for an external audit of its AI systems. The auditor requests evidence that the organization's AI security controls are effective. What is the most appropriate evidence to provide?

   • AA verbal assurance from the security team that controls are in place.
   • BLogs from continuous monitoring of AI system behavior and incident response records.
   • CA list of security tools purchased for AI monitoring.
   • DA written policy stating that AI security is a priority.
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A financial institution's AI trading system has been manipulated to execute unauthorized trades. The incident response team has been activated. What is the first step the team should take?

   • AAnalyze the trading logs to identify the root cause of the manipulation.
   • BIsolate the AI trading system from the market network to prevent further unauthorized trades.
   • CNotify all stakeholders, including regulators, before taking any action.
   • DShut down the entire trading platform, including manual trading systems.
   Reveal Solution Discussion Select an answer first