Updated Feb-2026 Test Engine to Practice AAISM Test Questions [Q89-Q103]

admin AAISM ISACA February 7, 2026

4.3/5 - (3 votes)

Updated Feb-2026 Test Engine to Practice AAISM Test Questions

AAISM Real Exam Questions Test Engine Dumps Training With 257 Questions

NEW QUESTION 89
Within an incident handling process, which of the following would BEST help restore end-user trust in an AI system?

Remediation of the AI system based on lessons learned

The AI model’s outputs are validated by team members

AI is used to monitor incident detection and alerts

The AI model prioritizes incidents based on business impact

NEW QUESTION 90
The PRIMARY goal of data poisoning attacks is to:

compromise the confidentiality of output data from the model

compromise the confidentiality of model input data

manipulate the behavior of the model during development

undermine the integrity of the AI system’s outputs

NEW QUESTION 91
An attack has occurred on an AI system that has been in use for two years. Which of the following would BEST mitigate the impact of the attack?

Monitoring AI systems for suspicious activities

Updating deployed training data with new adversarial data

Replacing the AI model with a new model that hides confidence levels

Implementing strict access controls to the model’s architecture

When an AI system experiences an attack after being in production for an extended period, the most effective mitigation strategy is to update the deployed training data with new adversarial data. This process strengthens the model’s resilience by retraining it to recognize and resist attack vectors that were previously unknown or unaccounted for. According to the AI Security Management™ (AAISM) framework, risk mitigation for AI systems must address model robustness through adversarial retraining, data quality improvement, and model lifecycle hardening rather than relying solely on reactive measures.
Why Option B is Correct:
* Incorporating adversarial examples into the training set enhances the system’s ability to correctly classify and withstand malicious inputs.
* This approach directly mitigates the vulnerability exploited in the attack and supports a proactive, continuous risk management cycle.
Why Other Options Are Incorrect:
* Option A: Monitoring helps detect suspicious activity but does not resolve the underlying vulnerability.
* Option C: Concealing confidence scores may reduce model transparency but does not address the attack mechanism or its root cause.
* Option D: Implementing access controls protects the model’s architecture but does not improve model robustness against input manipulation attacks.
Exact Extract from Official AAISM Study Guide:
“AI risk management requires continuous improvement following incidents. After an adversarial or data poisoning event, the preferred risk treatment involves retraining the model using adversarial data and updated datasets to enhance robustness. This ensures the AI model adapts to evolving threat landscapes rather than merely restricting access or obscuring outputs.” References:
AI Security Management™ (AAISM) Body of Knowledge: AI Risk Treatment and Mitigation Strategies, Adversarial Robustness and Resilience Engineering.
AI Security Management™ Study Guide: Model Lifecycle Security, Continuous Risk Treatment through Adversarial Retraining.
ISO/IEC 23894:2023, Clause 8.3.2 – Risk treatment through robustness improvement and adversarial data inclusion.

NEW QUESTION 92
An organization recently introduced a generative AI chatbot that can interact with users and answer their queries. Which of the following would BEST mitigate hallucination risk identified by the risk team?

Performing model testing and validation

Training the foundational model on large data sets

Ensuring model developers have been trained in AI risk

Fine-tuning the foundational model

NEW QUESTION 93
An organization is reviewing an AI application to determine whether it is still needed. Engineers have been asked to analyze the number of incorrect predictions against the total number of predictions made. Which of the following is this an example of?

Control self-assessment (CSA)

Model validation

Key performance indicator (KPI)

Explainable decision-making

NEW QUESTION 94
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?

Prompt injection

Jailbreaking

Remote code execution

Evasion

NEW QUESTION 95
An organization is designing an AI-based credit risk assessment system that will integrate with sensitive financial datasets. Which of the following would BEST support the implementation of security-by-design principles in the AI system’s architecture?

Segmenting AI services across containers to manage resource constraints

Restricting access to AI models using IP allow lists to reduce public exposure

Integrating differential privacy mechanisms into model training to limit data leakage

Applying threat modeling specific to AI components before deployment

NEW QUESTION 96
Which of the following is the MOST effective strategy for penetration testers assessing the security of an AI model against membership inference attacks?

Disabling AI model logging to reduce noise during testing

Measuring AI model accuracy on the test set

Analyzing AI model confidence scores to indicate training data

Generating synthetic data to replace the training data

NEW QUESTION 97
An organization has implemented a natural language processing model to respond to customer questions when personnel are not available. A pre-implementation security assessment revealed attackers could access sensitive company data through a chat interface injection attack. Which of the following is the BEST way to prevent this attack?

Ensuring continuous monitoring and data tagging

Manually reviewing AI model outputs

Implementing input validation and templates

Conducting regular information security audits

NEW QUESTION 98
An organization utilizes AI-enabled mapping software to plan routes for delivery drivers. A driver following the AI route drives the wrong way down a one-way street, despite numerous signs. Which of the following biases does this scenario demonstrate?

Selection

Reporting

Confirmation

Automation

NEW QUESTION 99
The PRIMARY purpose of adopting and implementing AI architecture within an organizational AI program is to:

Deploy fast and cost-efficient AI systems

Provide a basis for identifying threats and vulnerabilities

Align AI system components with business goals

Ensure powerful and scalable AI systems

NEW QUESTION 100
Which of the following is the MOST important consideration when an organization is adopting generative AI for personalized advertising?

Fraud risk

Reputational risk

Commercial risk

Regulatory risk

NEW QUESTION 101
Which of the following AI data management techniques involves creating validation and test data?

Training

Annotating

Splitting

Learning

NEW QUESTION 102
The PRIMARY purpose of adopting and implementing AI architecture as part of an organizational AI program is to:

ensure the development of powerful, efficient, and scalable AI systems

deploy fast and cost-efficient AI systems for rapidly changing environments

align the system components of AI with the business goals of the organization

provide a basis for identification of threats and vulnerabilities

NEW QUESTION 103
Which of the following factors is MOST important for preserving user confidence and trust in generative AI systems?

Bias minimization

Access controls and secure storage solutions

Transparent disclosure and informed consent

Data anonymization

ISACA AAISM Exam Syllabus Topics:

Topic	Details
Topic 1	AI Risk Management: This section of the exam measures the skills of AI Risk Managers and covers assessing enterprise threats, vulnerabilities, and supply chain risk associated with AI adoption, including risk treatment plans and vendor oversight.
Topic 2	AI Governance and Program Management: This section of the exam measures the abilities of AI Security Governance Professionals and focuses on advising stakeholders in implementing AI security through governance frameworks, policy creation, data lifecycle management, program development, and incident response protocols.
Topic 3	AI Technologies and Controls: This section of the exam measures the expertise of AI Security Architects and assesses knowledge in designing secure AI architecture and controls. It addresses privacy, ethical, and trust concerns, data management controls, monitoring mechanisms, and security control implementation tailored to AI systems.