Business Continuity and Incident Response AAISM Practice Questions (Page 4)

Questions 16–20

1. 16foundation · easy

   What is the purpose of implementing failover mechanisms for AI systems?

   • ATo reduce the size of the training dataset
   • BTo automatically switch to a redundant system when the primary system fails
   • CTo increase the speed of model training
   • DTo improve the accuracy of the AI model
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A hospital's AI-based patient triage system is critical for emergency department operations. The system relies on real-time patient data feeds and a cloud-hosted inference service. The hospital is conducting a business impact analysis (BIA) for this system. Which metric is MOST important to capture in the BIA to prioritize recovery efforts?

   • AThe number of data scientists who developed the model.
   • BMaximum tolerable downtime (MTD) for the triage system, measured in hours.
   • CThe cost of the cloud hosting subscription per month.
   • DThe model's accuracy score on the latest validation set.
   Reveal Solution Discussion Select an answer first
4. 18application · medium

   A financial services firm uses a machine learning model to approve credit card transactions in real time. During a routine security review, the model's behavior changes subtly: it starts approving a higher percentage of high-risk transactions. The change is traced to a compromised training data pipeline that injected malicious samples. The incident response team has been activated. Which action should be taken FIRST to align with the incident response lifecycle?

   • AEradicate the malicious samples from the training dataset and retrain the model immediately.
   • BContain the compromised pipeline by isolating it from the training environment and halting further data ingestion.
   • CRecover the model by rolling back to the last known good version from backup.
   • DConduct a lessons-learned review to document the root cause and improve future detection.
   Reveal Solution Discussion Select an answer first
5. 19application · medium

   A retail bank uses an AI chatbot for customer service. The chatbot is integrated with the bank's core banking system. The bank is conducting a business impact analysis (BIA) for the chatbot. Which finding would MOST likely justify a high recovery priority?

   • AThe chatbot handles 70% of customer inquiries, and its outage would significantly increase call center volume.
   • BThe chatbot was developed by a third-party vendor.
   • CThe chatbot uses a pre-trained language model.
   • DThe chatbot's training data is updated monthly.
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   A government agency has a cybersecurity incident response plan that follows the NIST framework. The agency is now deploying AI systems and needs to integrate AI incident response into the existing plan. Which approach BEST aligns with the NIST framework?

   • AIgnore the NIST framework and create a standalone AI incident response plan.
   • BUse the NIST framework only for non-AI incidents and handle AI incidents separately.
   • CReplace the NIST framework with a new AI-specific incident response framework.
   • DMap AI-specific incident types and response procedures to the existing NIST phases (Identify, Protect, Detect, Respond, Recover).
   Reveal Solution Discussion Select an answer first