Microsoft Certified: AI Business Professional (beta) Practice Exam 2025: Latest Questions

A retail manager wants to reduce customer wait times by forecasting the number of cashiers needed per hour using past store traffic and sales data. Which AI workload best fits this goal?

A company wants employees to ask natural-language questions about internal policy documents and get answers with citations to the source paragraphs. Which Azure approach is most appropriate?

A healthcare organization is building an AI assistant that must avoid revealing personal health information (PHI) in responses and must detect and block requests for sensitive data. Which feature best supports this requirement?

A business stakeholder asks for a way to measure whether a new customer-support chatbot is successful after launch. Which metric is the BEST business-aligned success measure?

An operations team wants to detect unusual spikes in IoT sensor readings that may indicate equipment failure. They have limited labeled failure data. Which solution is most appropriate?

A company plans to use generative AI to draft marketing emails. The legal team requires a process to review outputs for brand compliance and to record who approved final content. What is the BEST governance approach?

A team built a question-answering solution over internal documents. Users report that answers are fluent but occasionally incorrect and not supported by the documents. What is the MOST effective first step to reduce these errors?

A company wants to decide whether to build, buy, or customize an AI solution for invoice processing. They need a foundational approach that balances time-to-value with control over the solution. What should they do FIRST?

A bank is deploying an AI model to recommend credit limit increases. Regulators require the bank to provide understandable reasons for each recommendation and to demonstrate that outcomes are not discriminatory. Which combination BEST addresses these requirements?

An enterprise wants to roll out AI copilots across multiple departments. They must ensure consistent security controls, monitor usage, and prevent data leakage while allowing teams to experiment. Which operating model is MOST appropriate?

A retail manager wants a quick way to summarize customer feedback from surveys and identify the top themes. The solution must require minimal data science expertise and be deployable as an API. What should you recommend?

A business leader asks whether a proposed AI solution is best described as classification or regression. The solution predicts whether an insurance claim is likely to be fraudulent (Yes/No). Which type of AI task is this?

A company is rolling out an internal AI assistant that summarizes meeting notes. Leadership wants employees to know when content is AI-generated and how to report mistakes. Which action best supports responsible AI transparency?

A healthcare organization plans to analyze patient call recordings to extract key topics and sentiment. The organization must convert speech to text first and then perform text analytics. Which combination of Azure AI services is most appropriate?

A bank wants to deploy a generative AI feature that drafts customer email responses. The legal team requires that no customer personal data be used to train models without explicit approval, and they want controls to reduce the chance of the model outputting sensitive data. What should you recommend first?

A company wants to reduce time spent searching internal policies. They plan to build a chat experience that answers questions using their own documents, and they want the answers to cite the source policy sections. Which architecture pattern best fits this need?

An AI solution is performing well in pilot but degrades after several months in production because customer behavior changes. Which approach best addresses this issue?

A company uses Azure AI services to analyze documents. They discover that requests are failing intermittently due to throttling during peak hours. What is the best immediate mitigation?

A public-sector agency is deploying an AI model to assist with benefits eligibility screening. They must demonstrate that outcomes do not unfairly disadvantage a protected group. Which action best supports this requirement?

A company wants to deploy a customer-facing AI chatbot in multiple regions. Requirements include: low-latency responses, data residency controls, and a consistent safety policy for prompts and outputs. Which design is most appropriate?

A retail company wants to launch an AI feature that summarizes product reviews for shoppers. The business sponsor is concerned about inaccurate or misleading summaries impacting customer trust. What is the BEST first step to reduce business risk before broad release?

A customer support center wants to route incoming emails to the correct department (billing, technical support, returns, or account updates). They have thousands of labeled examples from the last year. Which Azure AI approach is MOST appropriate?

A bank plans to use an AI model to recommend credit limit increases. Regulators require that the bank can explain key factors influencing each recommendation to customers and auditors. What should the bank prioritize in its AI governance plan?

A manufacturing firm is building an AI solution to detect defects from images captured on a production line. During testing, the model performs well overall but fails more often for one factory location with different lighting and camera angles. What is the MOST likely cause and best next step?

A company wants to deploy a generative AI assistant for employees. Leadership requires that: (1) responses cite internal sources, (2) confidential documents must never be exposed to unauthorized staff, and (3) the system must be monitored for prompt injection attempts. Which design is BEST aligned to these requirements?