Isaca AAIA Dumps

Question # 1

An organization plans to implement an AI search and chatbot solution provided by an external vendor. Which of the following is MOST important for an IS auditor to confirm? 

A. Whether the solution is tied to a bilingual evaluation understudy (BLEU) metric 
B. Whether the vendor can provide an independent third-party attestation 
C. Whether the solution effectively enables web integration 
D. Whether the organization or the vendor has the test plan for the solution 

Show Answer

Question # 2

An organization using AI to create digital content faces challenges in protecting its intellectual property. Which of the following is the BEST way to mitigate this risk?

A. Separate training and production data. 
B. Implement watermarking.
C. Apply adversarial training. 
D. Implement a firewall. 

Show Answer

Question # 3

Which of the following pre-processing steps would MOST effectively justify an AI model's decision to a non-technical stakeholder?

A. Permutation feature importance 
B. Local interpretable model-agnostic explanations (LIME) 
C. Partial dependence plots 
D. AI model penetration testing 

Show Answer

Question # 4

Which of the following techniques BEST supports machine learning (ML) training in sentiment analysis?

A. Analysis of variance 
B. Log-rank test 
C. Image recognition 
D. Logistic regression 

Show Answer

Question # 5

An IS auditor reviewing the latest AI chatbot release identifies that, despite high accuracy rates, non-English users complain about the model's poor accuracy. Which of the following controls is BEST at ensuring detection of subgroup regressions?

A. Weighted metric with higher accuracy targets 
B. Human review for non-English languages after go-live 
C. Translate all outputs to English and evaluate in English for consistency 
D. Comparative language evaluations with parity thresholds 

Show Answer

Question # 6

Which of the following represents the PRIMARY benefit of reviewing model cards during AI model acquisition and risk assessment? 

A. Complying with documentation requirements 
B. Identifying model vulnerabilities and training needs 
C. Establishing legally binding contracts for model use 
D. Understanding model intent and performance context 

Show Answer

Question # 7

Which of the following is the MOST essential attribute of an AI-driven audit tool?

A. Explainability of model conclusions 
B. Optimization of audit resources 
C. Use of labeled training datasets 
D. Support for a range of statistical methods 

Show Answer

Question # 8

An IS auditor is reviewing a dataset related to a restaurant's food delivery process. Which of the following data preparation techniques should be used to ensure that no single feature dominates and all fields are weighted appropriately? 

A. Converting Postal Code values from text format to integers 
B. Inserting average values where Distance data is missing 
C. Removing duplicate entries for Customer Address field 
D. Scaling the Distance From Restaurant field 

Show Answer

Question # 9

Which of the following is the GREATEST risk when an organization uses publicly available social media data to train its AI model without obtaining explicit user consent?

A. Introduction of inaccurate and incomplete data 
B. Data with questionable labeling 
C. User distrust caused by lack of transparency in AI model training 
D. Violation of data privacy regulations leading to legal penalties 

Show Answer

Question # 10

When initiating an AI governance program, which of the following is MOST critical to ensure the AI system aligns with organizational objectives and stakeholder needs?

A. Defining AI regulatory compliance requirements before business needs 
B. Prioritizing AI hardware and infrastructure acquisition 
C. Establishing continuous AI system monitoring 
D. Identifying specific processes for the AI use case 

Show Answer

Question # 11

The internal audit department of a large law firm plans to implement an AI solution to review and analyze contracts stored in its enterprise resource planning (ERP) system. Which of the following AI technologies is MOST suitable for this requirement?

A. Deep learning 
B. Neural networks 
C. Natural language processing (NLP) 
D. Machine learning (ML) 

Show Answer

Question # 12

An IS auditor has discovered that an organization's AI system is less accurate than the required threshold and recommends management implement cross-referencing multiple models in order to produce more accurate results. This is an example of which type of risk response?

A. Avoidance 
B. Acceptance 
C. Transfer 
D. Mitigation 

Show Answer

Question # 13

Which of the following AI documents would support an IS auditor assessing hyperparameter tuning records?

A. Data sheets 
B. Model development logs 
C. Explainability reports 
D. Risk assessment reports 

Show Answer

Question # 14

From an audit perspective, which of the following BEST supports the objectives of an AI governance program?

A. Defining roles, responsibilities, and training for all stakeholders on AI ethics
B. Establishing an AI ethics committee or responsible AI review board 
C. Developing an AI risk register to record potential algorithmic bias 
D. Defining AI policies for model interpretability, transparency, and resiliency

Show Answer

Question # 15

An IS auditor learns that an organization uses AI facial recognition technology for security purposes. Which of the following is the GREATEST ethical concern with this practice? 

A. Difficulty of enforcing access controls to facial recognition data 
B. Potential bias against certain demographic groups 
C. Absence of global standards for the use of AI in biometric security 
D. Individuals beginning to interact more cautiously 

Show Answer

Question # 16

A financial institution's customer service chatbot sometimes gives users incorrect legal advice, leading to several customer complaints and potential regulatory exposure. Which of the following should be done FIRST to mitigate this risk? 

A. Expand the chatbot's training dataset. 
B. Retrain the chatbot on a more diverse dataset.
 C. Introduce a human-in-the-loop review process. 
D. Review chatbot technical documentation. 

Show Answer

Question # 17

Which of the following could be used to BEST identify underlying patterns in control effectiveness within unlabeled data elements?

A. XGBoost learning
 B. Reinforcement learning (RL) 
C. Unsupervised learning 
D. Random forest

Show Answer

Question # 18

Which of the following would be the BEST time to create a data governance plan for a new AI solution?

A. During the testing phase 
B. Prior to moving into production 
C. As part of the initial design 
D. During data collection 

Show Answer

Question # 19

Which of the following techniques would be MOST effective as part of incident management procedures for a prompt injection attack? 

A. Fine-tune the AI model. 
B. Scan inputs for code-like structure of text. 
C. Deploy input validation to sanitize abuse prompts. 
D. Monitor the prompts for excessive special characters. 

Show Answer

Question # 20

An AI team is developing a natural language processing (NLP) chatbot to assist customers. The collected data includes informal speech and multiple languages. Which pre-processing step is MOST important?

A. Standardize informal language using dictionaries and incorporate translation. 
B. Perform entity recognition to extract key information. 
C. Apply tokenization and stemming to reduce complexity. 
D. Remove all non-English data to avoid complexity. 

Show Answer