Medium — AB-731 Practice Quiz

Question 1

A retail organization is planning its first large-scale rollout of Microsoft 365 Copilot. During readiness assessment, the AI adoption team discovers that access permissions across SharePoint and OneDrive are inconsistent, with many files overshared to 'Everyone' groups. Leadership is eager to proceed but the team raises a concern that must be addressed before deployment. Which barrier to adoption does this finding most directly represent, and what is the primary risk if it is ignored?

Accepted Answer

A data governance and oversharing barrier — Copilot could surface sensitive content to users who should not have access, creating privacy and security exposure

Answer

A licensing barrier — the organization must purchase additional pay-as-you-go capacity before Copilot can index files

Answer

A change-management barrier — employees will resist using Copilot until they receive formal training on prompt writing

Answer

A model reliability barrier — Copilot may fabricate answers because the underlying files are inconsistently structured

Question 2

A retail company is finalizing its business case for a broad Microsoft 365 Copilot rollout to 5,000 employees. The finance team has only budgeted for the per-user Copilot license cost. As the AI Transformation Leader reviewing the plan, which additional cost impact is MOST commonly underestimated and should be explicitly added to the adoption budget?

Accepted Answer

The cost of change management activities such as training, an AI champions program, and ongoing enablement

Answer

The cost of purchasing new physical workstations for every employee

Answer

The cost of migrating the entire tenant to a different Microsoft 365 region

Answer

The cost of rebuilding all existing line-of-business applications from scratch

Question 3

A manufacturing company is launching a Microsoft 365 Copilot pilot with 200 employees across finance and operations. The AI transformation leader wants to prove the business value of the pilot before requesting funding for a broader rollout. Which action should the leader take FIRST when planning the adoption program?

Accepted Answer

Establish baseline productivity and outcome metrics before the pilot begins so post-pilot results can be measured against them

Answer

Purchase pay-as-you-go licenses for all 200 users to minimize upfront commitment during the pilot

Answer

Recruit AI champions from each department to promote adoption after the pilot concludes

Answer

Deploy Copilot to all users immediately and gather anecdotal feedback through informal surveys

Question 4

A retail company with 8,000 employees wants to deploy Microsoft 365 Copilot. The AI transformation leader is concerned about employee resistance, uneven data readiness across departments, and the need to demonstrate measurable value to justify continued investment to the executive board. Which adoption approach should the leader recommend?

Accepted Answer

Begin with a phased pilot in a few well-prepared departments, measure outcomes, refine the approach, and expand based on demonstrated value

Answer

Deploy Copilot to all 8,000 employees simultaneously to maximize the productivity impact and speed of return on investment

Answer

Delay any deployment until every department has fully remediated its data governance and permission issues across the entire tenant

Answer

Purchase licenses only for the executive board first so leadership can personally evaluate the tool before any wider rollout

Question 5

A retail company is preparing to deploy Microsoft 365 Copilot to 5,000 employees. During adoption planning, the IT security lead warns that Copilot surfaces content based on each user's existing SharePoint and OneDrive permissions, and that years of loosely governed file sharing means many documents are accessible far more broadly than intended. Which action should the adoption team prioritize BEFORE the broad rollout to address this privacy and oversharing risk?

Accepted Answer

Remediate access permissions and apply sensitivity labels so Copilot only surfaces content each user is genuinely authorized to see

Answer

Purchase additional Copilot licenses to ensure every employee can access the tool simultaneously

Answer

Fine-tune the underlying model on the company's document library to improve response relevance

Answer

Delay the entire program until the AI council drafts a new brand-voice prompt engineering standard

Question 6

A manufacturing company is launching its first enterprise-wide Copilot deployment. The CIO wants to establish a dedicated AI adoption team that can drive change management, measure business value, and ensure the rollout aligns with organizational goals. Which combination of roles should the CIO prioritize when forming this adoption team?

Accepted Answer

A cross-functional group including an executive sponsor, business unit leaders, change management specialists, and IT professionals

Answer

Only IT infrastructure engineers, since the deployment is fundamentally a technical provisioning task

Answer

Exclusively external consultants who specialize in AI so internal staff are not distracted from daily operations

Answer

A single AI project manager who reports usage metrics to leadership each quarter

Question 7

A manufacturing company is rolling out Microsoft 365 Copilot across its operations. Early pilot feedback shows that many frontline supervisors are reluctant to use the tool because they fear it will replace their jobs and they do not understand how it fits their daily tasks. The AI Transformation Leader is assembling an adoption team to address this. Which action should the adoption team prioritize to overcome this specific barrier to adoption?

Accepted Answer

Develop a change management and communication plan that clarifies how Copilot augments supervisors' work and includes role-specific training

Answer

Purchase additional Copilot licenses so every supervisor has guaranteed access before addressing their concerns

Answer

Restrict Copilot to a small technical team until the organization builds a full responsible AI governance framework

Answer

Switch the deployment from included Microsoft 365 licenses to a pay-as-you-go model to reduce upfront cost

Question 8

A retail company deploys a customer-facing chatbot built on generative AI to answer product questions and handle returns. During adoption planning, the AI council reviews the solution against Microsoft's responsible AI standard. Legal and customer experience teams raise concern that customers may not realize they are interacting with an AI rather than a human agent, and that the bot's suggestions could be mistaken for authoritative advice. Which responsible AI principle should the council apply to address this specific concern?

Accepted Answer

Transparency

Answer

Reliability and safety

Answer

Privacy and security

Answer

Fairness

Question 9

A manufacturing company has deployed Microsoft 365 Copilot to 2,000 employees. After three months, usage analytics show that adoption has stalled, with many employees unsure how Copilot applies to their daily tasks. The AI Transformation Leader wants a sustainable, peer-driven approach to increase confidence and everyday usage across departments. Which adoption initiative best addresses this situation?

Accepted Answer

Establish an AI champions program that identifies enthusiastic early adopters in each department to mentor colleagues and share practical use cases

Answer

Switch all users from the included Microsoft 365 Copilot license to a pay-as-you-go model to reduce cost until adoption improves

Answer

Create an AI council to define responsible AI governance principles and approve future AI projects

Answer

Deploy additional Foundry Tools with a commitment-tier subscription to expand the platform's technical capabilities

Question 10

A regional retail organization is launching a Microsoft 365 Copilot rollout across 40 stores. The adoption team wants to build an AI champions program to accelerate uptake. During planning, a store manager suggests that champions should be selected purely based on which employees have the deepest technical certifications in Azure and data engineering. The adoption lead disagrees. Which selection approach best reflects the purpose of an AI champions program?

Accepted Answer

Select enthusiastic, respected employees from across different roles and locations who can advocate for adoption, share practical use cases, and support peers

Answer

Choose champions primarily for their technical depth so they can troubleshoot backend configuration and API integrations for their peers

Answer

Assign the IT security team as the champions since they already control governance and access policies for AI tools

Answer

Limit champions to senior executives so leadership can mandate the use of Copilot from the top down

Question 11

A large insurance company is establishing an AI council as part of its responsible AI governance program. During the first meeting, members debate what the council's primary purpose should be. As the AI Transformation Leader, which statement best describes the appropriate role of the AI council in the organization's governance model?

Accepted Answer

To define enterprise-wide responsible AI policies, principles, and oversight standards that individual project teams then apply and are accountable to

Answer

To personally build and test every AI model before it is deployed so quality is guaranteed across all departments

Answer

To replace individual project teams' decision-making by directly approving every prompt, dataset, and configuration change

Answer

To serve as the help desk that resolves end-user support tickets for Microsoft 365 Copilot licensing issues

Question 12

A large financial services organization is establishing an AI council to oversee responsible AI adoption across the enterprise. The Chief Digital Officer wants the council to be effective at both setting policy and ensuring solutions meet responsible AI standards. Which approach should the organization take when forming the AI council?

Accepted Answer

Create a cross-functional council that includes leadership, legal, compliance, security, business unit representatives, and technical experts to provide diverse oversight

Answer

Staff the council exclusively with data scientists and machine learning engineers, since they best understand how the models work

Answer

Delegate all AI governance responsibilities to the existing IT change advisory board to avoid creating new committees

Answer

Limit the council to external consultants and auditors so decisions remain fully independent of internal business interests

Question 13

A retail company is evaluating a third-party AI recruiting tool that its AI council must approve before deployment. During review, the council discovers the vendor cannot explain how the model produces candidate rankings and provides no documentation on the training data or decision logic. The council wants to ensure the deployed solution meets Microsoft's responsible AI standards. Which responsible AI principle is most directly at risk given the vendor's inability to explain the model's outputs?

Accepted Answer

Transparency

Answer

Reliability and safety

Answer

Privacy and security

Answer

Inclusiveness

Question 14

A finance operations team regularly receives raw CSV exports from multiple regional sales systems. They need Microsoft 365 Copilot to combine these files, run calculations, identify revenue trends across quarters, and produce a clear data-driven summary with charts. They want the tool to actually process the data programmatically rather than just retrieve documents or browse the web. Which Microsoft 365 Copilot agent should the team use?

Accepted Answer

The Analyst agent, which uses code execution to reason over and transform the provided data into insights

Answer

The Researcher agent, which performs multi-step research across internal and web sources

Answer

Microsoft Copilot Studio, to build a topic-based conversational chatbot for the finance portal

Answer

Microsoft 365 Copilot Chat on mobile, to summarize the CSV files from a phone

Question 15

A regional sales director uploads three quarters of raw sales data as CSV files to Microsoft 365 Copilot and asks for revenue trends, correlations between discount levels and deal size, and a chart summarizing the findings. The director needs the analysis performed directly on the uploaded numeric data with reproducible calculations. Which capability should the director use?

Accepted Answer

The Analyst agent, which runs Python code to process the uploaded data and produce calculated insights and visualizations

Answer

The Researcher agent, which performs multi-step reasoning across web and tenant sources to compile a report

Answer

Microsoft 365 Copilot Chat on the web, which answers grounded questions without executing analytical code

Answer

Microsoft Copilot Studio, to build a custom topic-based agent for parsing CSV files

Question 16

A retail company is building a customer-support chatbot on Microsoft Foundry. The support content lives in thousands of PDF manuals and internal wiki pages. The team wants the chatbot's generative model to return answers grounded in this content, with the ability to retrieve the most semantically relevant passages at query time. Which Foundry-related service should the team use to index and retrieve this content for grounding?

Accepted Answer

Azure AI Search

Answer

Azure AI Vision

Answer

Microsoft Graph

Answer

Azure Machine Learning designer

Question 17

A retail company is building a RAG-based product support assistant on Microsoft Foundry. During testing, users report that keyword searches return technically relevant documents but often miss articles that describe the same problem using different wording (e.g., 'device won't power on' vs. 'unit fails to boot'). The team wants the retrieval layer in Azure AI Search to match on meaning rather than exact terms. Which Azure AI Search capability should they enable to address this?

Accepted Answer

Vector (embeddings-based) search with semantic ranking

Answer

Synonym maps applied to the keyword index

Answer

Scoring profiles that boost recently updated documents

Answer

Indexer scheduling to refresh content more frequently

Question 18

A logistics company receives thousands of scanned paper delivery receipts and photographed shipping labels each day. They want to automatically detect text, read printed and handwritten content, and identify objects within these images to feed structured data into their tracking system. The team wants a prebuilt capability within Microsoft Foundry rather than training a custom model from scratch. Which Foundry Tools capability best fits this need?

Accepted Answer

Azure AI Vision in Foundry Tools

Answer

Azure AI Search

Answer

Microsoft Graph connectors

Answer

Copilot Studio topics

Question 19

A logistics company already uses Microsoft 365 Copilot broadly. Their employees want Copilot to answer questions using real-time shipment status data that lives in a proprietary internal tracking system with a REST API. The company does not want to build a standalone chatbot; they want the answers to surface directly within the existing Microsoft 365 Copilot experience. Which approach best fits this need?

Accepted Answer

Extend Microsoft 365 Copilot by building a plugin/connector using the Microsoft 365 Copilot extensibility framework to reach the tracking system's API

Answer

Buy a third-party logistics SaaS chatbot and deploy it as a separate application outside Microsoft 365

Answer

Build a completely new custom large language model trained on the company's shipment data

Answer

Replace Microsoft 365 Copilot with a standalone Copilot Studio agent hosted only on the public website

Question 20

A financial analyst at a mid-sized manufacturing company spends hours each week analyzing quarterly sales data stored in large Excel workbooks. She wants to use Microsoft 365 Copilot to surface trends, generate formula-based calculations, and create summary charts directly within the spreadsheet she is working in. Which Copilot experience best matches this need?

Accepted Answer

Microsoft 365 Copilot embedded in Excel

Answer

Microsoft 365 Copilot Chat on the web

Answer

A custom agent built in Microsoft Copilot Studio

Answer

The Researcher agent in Microsoft 365 Copilot

Question 21

A 4,000-person professional services firm plans to roll out Microsoft 365 Copilot to 500 knowledge workers who will use it daily for drafting, summarization, and meeting recaps throughout their standard workday. Finance wants the most cost-predictable licensing model for this consistent, per-user daily usage. Which Copilot licensing approach should the AI transformation leader recommend?

Accepted Answer

The Microsoft 365 Copilot per-user monthly (or annual) subscription assigned to the 500 workers

Answer

The pay-as-you-go metered model billed by message/agent consumption for all 500 workers

Answer

The free Microsoft 365 Copilot Chat entitlement included with existing Microsoft 365 licenses

Answer

Azure AI Foundry Tools commitment-tier billing purchased for the 500 workers

Question 22

A sales director spends the first hour of every workday manually reading through dozens of overnight emails, deciding which require urgent responses and drafting replies to the most important ones. She has a Microsoft 365 Copilot license and wants to reduce this triage time without leaving her email client. Which Microsoft 365 Copilot capability within Outlook is the best fit for this business process?

Accepted Answer

Use Copilot in Outlook to summarize long email threads and draft context-aware replies directly in the inbox

Answer

Build a custom agent in Copilot Studio that reads her mailbox and posts summaries to a Teams channel

Answer

Use the Analyst agent to run Python analysis over exported email data and rank messages by priority

Answer

Use Microsoft Foundry to fine-tune a model on her past email responses for automated sending

Question 23

A retail organization has built a Copilot Studio customer-service agent that will handle a large, unpredictable spike in interactions during the holiday shopping season but only minimal traffic the rest of the year. Leadership wants to avoid committing to fixed monthly costs for capacity they will not consistently use. Which licensing approach best fits this consumption pattern?

Accepted Answer

Use the Copilot Studio pay-as-you-go (metered) billing model based on message consumption

Answer

Purchase Microsoft 365 Copilot seats for every customer-service employee

Answer

Buy a fixed number of prepaid monthly Copilot Studio message packs regardless of demand

Answer

Adopt an Azure AI Foundry Tools commitment tier for the agent workload

Question 24

A logistics company wants to build a conversational agent that lets warehouse staff check inventory and, when stock is low, automatically trigger a reorder by calling the company's existing procurement REST API. The team wants a low-code environment where non-developers can design the conversation flow and connect the API call. Which Microsoft capability best fits this requirement?

Accepted Answer

Microsoft Copilot Studio, using topics and actions to orchestrate the conversation and call the procurement API

Answer

Microsoft 365 Copilot Chat on the web, relying on its built-in web grounding

Answer

The Researcher agent in Microsoft 365 Copilot

Answer

Azure AI Search with a custom vector index over inventory documents

Question 25

A marketing team already has Microsoft 365 Copilot licenses. They want a lightweight, purpose-built agent that surfaces answers from their SharePoint campaign guidelines and a specific set of Word playbooks directly inside the Microsoft 365 Copilot chat experience. They have no developers and want the simplest way to create and share it with the team. Which approach best fits this need?

Accepted Answer

Build a declarative agent using the agent builder within Microsoft 365 Copilot, grounding it in the selected SharePoint sites and files

Answer

Provision an Azure AI Foundry Agent Service project and deploy a custom orchestrated agent with a REST endpoint

Answer

Create an Azure AI Search index and build a standalone custom RAG web application for the team

Answer

Fine-tune a foundation model from the Microsoft Foundry model catalog on the campaign guidelines