Isaca AAISM Online Practice Questions and Exam Preparation

Isaca AAISM Online Questions & Answers

• Question 131:

  An attacker crafts inputs to a large language model (LLM) to exploit output integrity controls. Which of the following types of attacks is this an example of?

  A. Prompt injection
  B. Jailbreaking
  C. Remote code execution
  D. Evasion
  A. Prompt injection

  ---

  According to the AAISM framework, prompt injection is the act of deliberately crafting malicious or manipulative inputs to override, bypass, or exploit the model's intended controls. In this case, the attacker is targeting the integrity of the model's outputs by exploiting weaknesses in how it interprets and processes prompts. Jailbreaking is a subtype of prompt injection specifically designed to override safety restrictions, while evasion attacks target classification boundaries in other ML contexts, and remote code

  execution refers to system-level exploitation outside of the AI inference context. The most accurate classification of this attack is prompt injection.

  References:

  AAISM Exam Content Outline - AI Technologies and Controls (Prompt Security and Input Manipulation)

  AI Security Management Study Guide - Threats to Output Integrity

• Question 132:

  Which of the following is the MOST important consideration when an organization is adopting generative AI for personalized advertising?

  A. Fraud risk
  B. Reputational risk
  C. Commercial risk
  D. Regulatory risk
  D. Regulatory risk

  ---

  In AAISM, usage of AI for activities involving personal data and profiling, such as personalized advertising, is explicitly mapped to stringent regulatory and compliance requirements (e.g., data protection, consent, profiling limitations, fairness obligations). The material notes that these activities may trigger "heightened regulatory scrutiny, mandatory impact assessments, and potential penalties for non-compliance." While reputational (B), fraud (A), and commercial (C) risks are all relevant, the primary, non-optional constraint is compliance with applicable regulations governing personal data, automated profiling, and targeted content. Failure in this area can lead not only to reputational harm but also to legal sanctions, enforced remediation, and operational restrictions. Therefore, regulatory risk is identified as the most important consideration when deploying generative AI for personalized advertising.

  References: AI Security Management (AAISM) Study Guide - AI, Privacy, and Regulatory Compliance; High-Risk Use Cases and Profiling.

• Question 133:

  In the context of generative AI, which of the following would be the MOST likely goal of penetration testing during a red-teaming exercise?

  A. Generate outputs that are unexpected using adversarial inputs
  B. Stress test the model's decision-making process
  C. Degrade the model's performance for existing use cases
  D. Replace the model's outputs with entirely random content
  A. Generate outputs that are unexpected using adversarial inputs

  ---

  AAISM's risk management content describes red-teaming in generative AI as focused on deliberately crafting adversarial prompts to test whether the model produces unexpected or undesired outputs that violate safety, integrity, or compliance standards. The goal is not to stress system performance or randomly disrupt outputs, but rather to uncover vulnerabilities in how the model responds to manipulative inputs. This allows organizations to improve resilience against prompt injection, jailbreaking, or harmful

  content generation. The correct answer is therefore generate outputs that are unexpected using adversarial inputs.

  References:

  AAISM Exam Content Outline - AI Risk Management (Red-Team Testing and Adversarial Exercises)

  AI Security Management Study Guide - Penetration Testing in Generative AI Contexts

• Question 134:

  During red-team testing of an AI system used for lending decisions, which technique BEST simulates a data poisoning attack?

  A. Adding noise to output predictions
  B. Stealing model weights
  C. Inputting encrypted data
  D. Corrupting training datasets to manipulate outcomes
  D. Corrupting training datasets to manipulate outcomes

  ---

  AAISM defines data poisoning as intentional manipulation of the training data to influence model behavior or outputs. Corrupting training data (D) is the exact definition of this attack type. Noise injection (A) is model degradation testing. Model theft (B) is exfiltration. Encrypted data (C) is irrelevant.

  References: AAISM Study Guide - AI Threats; Data Poisoning Attacks.

• Question 135:

  An organization plans to implement a new AI system. Which of the following is the MOST important factor in determining the level of risk monitoring activities required?

  A. The organization's risk appetite
  B. The organization's number of AI system users
  C. The organization's risk tolerance
  D. The organization's compensating controls
  C. The organization's risk tolerance

  ---

  AAISM risk management guidance clarifies that the organization's risk tolerance is the most important factor in determining how much monitoring is needed. Risk tolerance specifies the amount of risk the organization is willing to accept and defines the threshold for triggering monitoring or mitigation activities. Risk appetite is broader and strategic, while tolerance sets the operational limits. The number of users may influence scale, and compensating controls may affect resilience, but neither dictates

  monitoring intensity as directly as risk tolerance.

  References:

  AAISM Study Guide - AI Risk Management (Risk Appetite vs. Tolerance)

  ISACA AI Security Management - Monitoring Based on Risk Tolerance

• Question 136:

  Which of the following BEST strengthens information security controls around the use of generative AI applications?

  A. Ensuring controls exceed industry benchmarks
  B. Monitoring AI outputs against policy
  C. Implementing a kill switch
  D. Validating AI model training data
  B. Monitoring AI outputs against policy

  ---

  AAISM identifies continuous monitoring of AI outputs-especially generative outputs-as the most effective security control, ensuring that violations, unsafe responses, data leakage, and policy-breaking behavior are detected and corrected. A kill switch (C) is a last-resort measure, not a primary control. Exceeding benchmarks (A) does not ensure relevance. Validating training data (D) is important but insufficient for generative output risks.

  References: AAISM Study Guide - Generative AI Security Controls; Output Monitoring and Policy Alignment.

• Question 137:

  Which of the following is the MAIN objective of the operational phase of AI life cycle management?

  A. Monitor model performance
  B. Align the model to business needs
  C. Optimize the model's algorithms
  D. Obtain end-user feedback on the model
  A. Monitor model performance

  ---

  In the operational phase, AAISM emphasizes continuous monitoring of models for performance, stability, robustness, drift, data quality, security events, and policy compliance. This includes telemetry, thresholds, alerts, incident response for AI failures, and evidence collection for audits. Alignment to business needs is established earlier in planning/governance; algorithmic optimization and feedback collection are supporting activities, but the primary operational objective is live monitoring and assurance to keep risk within tolerance. References:. AI Security Management (AAISM) Body of Knowledge: AI Life Cycle-Operate/Monitor; Ongoing Performance and Drift Monitoring; AI Incident Management. AAISM Study Guide: Operational Controls, Metrics and SLAs/SLOs; Evidence and Audit Readiness in Production

• Question 138:

  Which of the following is MOST important to consider when validating a third-party AI tool?

  A. Terms and conditions
  B. Right to audit
  C. Industry analysis and certifications
  D. Roundtable testing
  B. Right to audit

  ---

  The AAISM framework specifies that when adopting third-party AI tools, the right to audit is the most critical contractual and governance safeguard. This ensures that the organization can independently verify compliance with security, privacy, and ethical requirements throughout the lifecycle of the tool. Terms and conditions provide general usage guidance but often limit liability rather than ensuring transparency. Industry certifications may indicate good practice but do not substitute for direct verification.

  Roundtable testing is useful for evaluation but lacks enforceability. Only the contractual right to audit provides formal assurance that the tool operates in accordance with organizational policies and external regulations.

  References:

  AAISM Exam Content Outline - AI Governance and Program Management (Third-Party Governance)

  AI Security Management Study Guide - Vendor Oversight and Audit Rights

• Question 139:

  Which of the following is the MOST effective way to identify and address security risk in an AI model?

  A. Assign staff to review AI model outputs for accuracy
  B. Conduct threat modeling to identify vulnerabilities and possible attack methods
  C. Encrypt the training data and model parameters to prevent unauthorized access
  D. Add more data to the model to increase its accuracy and reduce errors
  B. Conduct threat modeling to identify vulnerabilities and possible attack methods

  ---

  AI/ML threat modeling is the most effective structured method to both identify and address model security risks. It systematically surfaces attack classes (poisoning, evasion, membership inference, model extraction, inversion), maps system-specific attack surfaces (data pipelines, feature stores, training artifacts, inference APIs), and drives prioritized mitigations (ingestion validation, robust training, rate-limiting, watermarking, differential privacy, monitoring, red teaming). Output spot-checking (A) finds errors but not security vulnerabilities; encryption (C) protects confidentiality but does not reveal threats or mitigate inference-time attacks; adding data (D) may improve accuracy but does not target adversarial risk.

  References: AI Security Management (AAISM) Body of Knowledge - AI Risk Identification and Threat Modeling; Attack Surface Analysis for ML; Risk Treatment Planning. AAISM Study Guide - Evasion / Poisoning/Extraction Controls; Mapping Risks to Controls; Validation and Assurance Activities.

• Question 140:

  Which of the following should be included in an AI acceptable use policy?

  A. AI training data requirements
  B. Data collection and storage processes
  C. Ethical and legal compliance standards
  D. AI monitoring requirements
  C. Ethical and legal compliance standards

  ---

  An AI acceptable use policy (AUP) sets the organizational expectations and boundaries for how AI systems may be used by employees and third parties. AAISM guidance places emphasis on ethical and legal compliance standards as core elements of an AUP to govern responsible behavior, prevent misuse, and align with regulatory and organizational principles. While data requirements, collection/storage processes, and monitoring may be covered in adjacent standards and procedures (e.g., data management policies, SOPs, and operational runbooks), the AUP's essential function is to codify permissible use anchored to ethics, legality, and organizational values.

  References: AI Security Management (AAISM) Body of Knowledge - AI Governance Policies and Codes of Conduct; Responsible Use Principles. AAISM Study Guide - Policy Hierarchy and Control Mapping; Acceptable Use and Staff Obligations.