Comprehend Artificial Intelligence (AI)/Machine Learning (ML) CCSP Practice Questions (Page 2)

Questions 6–10

1. 6application · medium

   A security team uses an AI/ML model to classify network traffic as malicious or benign. The model is trained on a dataset that includes labeled examples of known malware traffic. Recently, the team suspects that an attacker may have injected malicious samples into the training data to cause the model to misclassify certain attacks. Which technique is most effective to prevent this type of model poisoning?

   • AImplement strict data validation and provenance checks on all training inputs.
   • BUse a diverse set of data sources to reduce the impact of any single source.
   • CRegularly retrain the model with new data to keep it up to date.
   • DIncrease the model's complexity to make it more resistant to manipulation.
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   How does AI/ML enhance SOAR platforms in cloud security?

   • ABy eliminating all false positives
   • BBy providing a user interface for analysts
   • CBy automating threat detection, analysis, and response actions
   • DBy replacing the need for any human oversight
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   In the context of AI/ML threat analysis, what does 'prioritization' of security alerts primarily achieve?

   • AIt assigns alerts to the least experienced analyst
   • BIt deletes low-priority alerts automatically to reduce storage
   • CIt converts all alerts into a single unified format
   • DIt ranks alerts based on potential impact and likelihood, guiding response efforts
   Reveal Solution Discussion Select an answer first
5. 9application · medium

   A company is training an AI/ML model to detect insider threats using historical employee activity logs. The data is collected from multiple sources, including HR systems, VPN logs, and file access records. The security team is concerned about the integrity of the training data. Which measure is most important to ensure the model is not trained on tampered or unreliable data?

   • AUse the most recent data available to ensure the model reflects current behavior.
   • BValidate the authenticity and completeness of each data source before using it for training.
   • CRely on the cloud provider's default data storage security.
   • DIncrease the volume of data to improve the model's accuracy.
   Reveal Solution Discussion Select an answer first
6. 10expert · hard

   A large enterprise uses an AI/ML-based anomaly detection system to monitor network traffic. The system is trained on historical traffic patterns and flags deviations from the baseline. Recently, the system has been generating an increasing number of false positives, causing alert fatigue among analysts. The security team suspects that the training data may have been contaminated with malicious traffic samples. Which combination of actions is most effective to address this issue?

   • ASwitch to a simpler rule-based detection system to avoid AI/ML complexity.
   • BValidate the integrity of the training data and implement continuous monitoring of model performance.
   • CRetrain the model with only the most recent traffic data and increase the anomaly threshold.
   • DIncrease the model's sensitivity to catch more anomalies, even if it means more false positives.
   Reveal Solution Discussion Select an answer first