A healthcare AI system that predicts patient deterioration has been compromised. The attack appears to have altered the model's decision boundaries, causing it to miss critical alerts. The incident response team needs to contain the incident. What is the most immediate containment step?

Isolate the AI system from the production network and switch to a manual review process.

Notify the regulatory authority before taking any action.

Delete the model and all associated logs to prevent further spread.

Immediately retrain the model with new data to restore its performance.

Why might incident response for an AI system require specialized procedures beyond traditional IT incident response?

Because AI incidents are always more severe than traditional incidents

Because AI systems do not require security monitoring

Because AI systems are always cloud-based

Because AI incidents may involve model-specific issues like data poisoning or adversarial attacks

A company's AI system for fraud detection has been generating an increasing number of false positives. The security team wants to set up monitoring to detect this issue early. Which monitoring metric is most appropriate for this scenario?

CPU utilization of the serving infrastructure

Model's training loss on the original dataset

Precision and recall of the model

Number of user complaints received

ISACAAdvanced in AI Security Management

Domain 3Objective 5

Security Controls and Monitoring AAISM Practice Questions (Page 2)

Part of the AI Technologies and Controls domain, which accounts for 38% of the AAISM exam.

29questions here

6free pages

5concepts

38%of the exam

Questions 6–10

6application · medium
A healthcare AI system that predicts patient deterioration has been compromised. The attack appears to have altered the model's decision boundaries, causing it to miss critical alerts. The incident response team needs to contain the incident. What is the most immediate containment step?
Select an answer first
7foundation · easy
Why might incident response for an AI system require specialized procedures beyond traditional IT incident response?
Select an answer first
8foundation · easy
What is the first step in an incident response procedure tailored to an AI security incident?
Select an answer first
9application · medium
A company's AI system for fraud detection has been generating an increasing number of false positives. The security team wants to set up monitoring to detect this issue early. Which monitoring metric is most appropriate for this scenario?
Select an answer first
10application · medium
A company is implementing AI security controls and needs to align them with industry best practices. The company's compliance officer recommends adopting a framework that is internationally recognized and covers AI management systems. Which framework should the company adopt?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to AAISM

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by ISACA. “AAISM” is a trademark of its owner, used for identification only.

Security Controls and Monitoring AAISM Practice Questions (Page 2)

Questions 6–10

A healthcare AI system that predicts patient deterioration has been compromised. The attack appears to have altered the model's decision boundaries, causing it to miss critical alerts. The incident response team needs to contain the incident. What is the most immediate containment step?

Why might incident response for an AI system require specialized procedures beyond traditional IT incident response?

What is the first step in an incident response procedure tailored to an AI security incident?

A company's AI system for fraud detection has been generating an increasing number of false positives. The security team wants to set up monitoring to detect this issue early. Which monitoring metric is most appropriate for this scenario?

A company is implementing AI security controls and needs to align them with industry best practices. The company's compliance officer recommends adopting a framework that is internationally recognized and covers AI management systems. Which framework should the company adopt?

Keep studying — AI Technologies and Controls