IBM A1000-050 - Assessment: Foundations of AI Practice Exam 2025: Latest Questions

A business stakeholder asks what makes a system “AI” instead of traditional software. Which description best matches an AI system?

A team is building a model to classify emails as spam or not spam using a historical dataset where each email is labeled. What type of machine learning is this?

Which use case is best suited for a conversational AI assistant?

An AI team wants to reduce privacy risk when training on customer data. Which action is a best practice?

A retailer wants to group customers into segments based on purchasing behavior without pre-defined labels. Which approach is most appropriate?

A model performs very well on the training set but poorly on new, unseen data. What is the most likely issue?

A bank plans to use an AI model to help approve loans. Regulators require the bank to provide reasons for adverse decisions to applicants. Which model characteristic is most important?

A company is deploying an AI solution for customer support. They want a design that reduces risk and improves reliability. Which architecture best practice should they apply?

A medical AI model shows high overall accuracy, but it misses a significant number of positive cases for a rare disease. The business impact of missing a positive case is high. Which metric should the team prioritize improving?

A company discovers its hiring-screening model disadvantages a protected group. They want to address this while maintaining governance controls. What is the best next step?

A help-desk team wants to automate answering common employee questions (e.g., password reset steps, VPN setup) using a conversational interface. The team wants the system to retrieve approved answers from an internal knowledge base rather than invent responses. Which approach is most appropriate?

A product manager asks whether a proposed AI model is doing classification or regression. The model outputs one of five possible defect categories for a manufactured part. What type of problem is this?

A data scientist wants to evaluate a binary classifier for loan default prediction. Defaults are rare, and missing a true default is costly. Which metric is generally more informative than raw accuracy in this situation?

A team trained a model and achieved very high performance on the training set but significantly lower performance on a validation set. What is the most likely issue?

A retailer is using an AI model to recommend products. They want recommendations to adapt based on recent user behavior (e.g., a user recently searched for running shoes). Which data characteristic is most important to incorporate?

A bank is deploying an AI model that influences credit decisions. Which action best supports responsible AI governance before and during deployment?

A team is building a computer vision system to detect safety helmet usage in a factory. They have 50,000 labeled images and want to reduce training time while maintaining good accuracy. Which approach is most appropriate?

An organization is concerned that its training dataset under-represents a protected group, which could lead to unfair outcomes. What is the most appropriate first step to address this risk?

A healthcare organization trains a model for patient readmission prediction using historical data. After deployment, hospital policies change and the model’s performance steadily degrades. What is the most likely explanation?

A team is fine-tuning a large language model for internal document summarization. They notice the model sometimes includes confidential customer details that appear in the training data. Which mitigation is most appropriate to reduce the risk of sensitive data being reproduced?

A retail team wants to group customers into segments based on purchasing behavior, but they do not have labeled data indicating which segment each customer belongs to. Which machine learning approach best fits this need?

A bank deploys a credit approval model and later discovers that approval rates differ significantly across protected groups, even when applicants have similar financial profiles. What is the most appropriate FIRST step to address this issue?

A customer support team wants an AI assistant to answer questions using only the company’s internal policy documents and to reduce the risk of the model inventing answers. Which architecture pattern best addresses this requirement?

A data scientist reports that a model’s performance looks excellent during development but drops sharply after deployment. Monitoring shows that customer behavior has changed due to a new marketing campaign, shifting the distribution of key input features. What is the most likely cause?

An organization wants to operationalize AI responsibly. Which combination of controls BEST supports accountability and auditability for an AI model in production?