AIGP Free Practice Test

- (Topic 1)
CASE STUDY
Please use the following answer the next question:
ABC Corp, is a leading insurance provider offering a range of coverage options to individuals. ABC has decided to utilize artificial intelligence to streamline and improve its customer acquisition and underwriting process, including the accuracy and efficiency of pricing policies.
ABC has engaged a cloud provider to utilize and fine-tune its pre-trained, general purpose large language model (“LLM”). In particular, ABC intends to use its historical customer data—including applications, policies, and claims—and proprietary pricing and risk strategies to provide an initial qualification assessment of potential customers, which would then be routed a human underwriter for final review.
ABC and the cloud provider have completed training and testing the LLM, performed a readiness assessment, and made the decision to deploy the LLM into production. ABC has designated an internal compliance team to monitor the model during the first month, specifically to evaluate the accuracy, fairness, and reliability of its output. After the first month in production, ABC realizes that the LLM declines a higher percentage of women's loan applications due primarily to women historically receiving lower salaries than men.
What is the best strategy to mitigate the bias uncovered in the loan applications?

1. A. Retrain the model with data that reflects demographic parity.
2. B. Procure a third-party statistical bias assessment tool.
3. C. Document all instances of bias in the data set.
4. D. Delete all gender-based data in the data set.

Correct Answer: A
Retraining the model with data that reflects demographic parity is the best strategy to mitigate the bias uncovered in the loan applications. This approach addresses the root cause of the bias by ensuring that the training data is representative and balanced, leading to more equitable decision-making by the AI model.
Reference: The AIGP Body of Knowledge stresses the importance of using high-quality,
unbiased training data to develop fair and reliable AI systems. Retraining the model with balanced data helps correct biases that arise from historical inequalities, ensuring that the AI system makes decisions based on equitable criteria.

- (Topic 2)
In the machine learning context, feature engineering is the process of?

1. A. Converting raw data into clean data.
2. B. Creating learning schema for a model apply.
3. C. Developing guidelines to train and test a model.
4. D. Extracting attributes and variables from raw data.

Correct Answer: D
In the machine learning context, feature engineering is the process of extracting attributes and variables from raw data to make it suitable for training an AI model. This step is crucial as it transforms raw data into meaningful features that can improve the model's accuracy and performance. Feature engineering involves selecting, modifying, and creating new features that help the model learn more effectively. Reference: AIGP Body of Knowledge on AI Model Development and Feature Engineering.

- (Topic 1)
According to the GDPR, what is an effective control to prevent a determination based solely on automated decision-making?

1. A. Provide a just-in-time notice about the automated decision-making logic.
2. B. Define suitable measures to safeguard personal data.
3. C. Provide a right to review automated decision.
4. D. Establish a human-in-the-loop procedure.

Correct Answer: D
The GDPR requires that individuals have the right to not be subject to decisions based solely on automated processing, including profiling, unless specific exceptions apply. One effective control is to establish a human-in-the-loop procedure (D), ensuring human oversight and the ability to contest decisions. This goes beyond just-in- time notices (A), data safeguarding (B), or review rights (C), providing a more robust mechanism to protect individuals' rights.

- (Topic 1)
Which of the following is NOT a common type of machine learning?

1. A. Deep learning.
2. B. Cognitive learning.
3. C. Unsupervised learning.
4. D. Reinforcement learning.

Correct Answer: B
The common types of machine learning include supervised learning, unsupervised learning, reinforcement learning, and deep learning. Cognitive learning is not a type of machine learning; rather, it is a term often associated with the broader field of cognitive science and psychology. Reference: AIGP BODY OF KNOWLEDGE and standard AI/ML literature.

- (Topic 2)
During the planning and design phases of the Al development life cycle, bias can be reduced by all of the following EXCEPT?

1. A. Stakeholder involvement.
2. B. Feature selection.
3. C. Human oversight.
4. D. Data collection.

Correct Answer: B
Bias in AI can be reduced during the planning and design phases through stakeholder involvement, human oversight, and careful data collection. While feature selection is critical in the development phase, it does not specifically occur during planning and design. Ensuring diverse stakeholder involvement and human oversight helps identify and mitigate potential biases early, and data collection ensures a representative dataset. Reference: AIGP Body of Knowledge on AI Development Lifecycle and Bias Mitigation.