AB-730 study plan: pass in 4 weeks

Four weeks at three to five hours a week is a comfortable timeline for AB-730 if you already use Microsoft 365. Spend week 1 on generative AI fundamentals, week 2 on prompts and chats, week 3 on content and collaboration, and week 4 on agents and Cowork plus a full review. This plan follows the outline in force from October 20, 2026.

Before you start

Read Microsoft’s study guide once, end to end. The bullet lists under each objective show the level of detail the exam expects.
Get hands-on access to Microsoft 365 Copilot if you can. Many questions describe a button or a feature by name, and ten minutes in the product beats an hour of reading. If you have no licence, use the free practice assessment and the exam sandbox to see how questions are phrased.
Keep a list of look-alike pairs and add to it every session: Copilot Chat versus an agent, a prebuilt versus a custom agent, a notebook versus a page, saving a prompt versus scheduling it.

Week 1: generative AI fundamentals (25–30%)

Goal: explain how Copilot answers, and where it can go wrong.

How Copilot keeps your organisation’s data private and secure, and how data protection limits what a prompt can return.
How context shapes a response: the data available to Copilot about your work, web data, and the app you are in.
When to use Copilot Chat, an agent or Cowork, and how features differ between Copilot experiences and apps.
Risks: inaccurate answers, prompt injection and over-reliance, and the checks that fit each task, such as verifying citations and human review.
Recognising and reducing risk to sensitive data.

Task: write one realistic work prompt, then list three ways its answer could be wrong and the check you would use for each.

Week 2: prompts and chats (20–25%)

Goal: write, reuse and organise prompts like a daily user.

What makes a prompt effective, including model selection and choosing which files or sources to reference.
Saving, scheduling and sharing a prompt, and refining one after a weak response.
How memory and instructions affect responses.
Finding, renaming and deleting a past chat, and organising and sharing information with notebooks.

Task: take a prompt that gives a mediocre result and improve it three times, changing one element each time (role, source, format). Note which change helped most.

Week 3: content and collaboration (20–25%)

Goal: produce and check business content across apps and meetings.

Connecting to sources beyond Microsoft 365 with connectors, and generating or analysing content with Work IQ and those connectors.
Generating images, visuals and charts, and working across Word, Excel and PowerPoint.
Checking that AI-generated content meets the business goal.
Meetings: preparing for one, the Facilitator agent, real-time catchup, intelligent recap, and Copilot Pages for shared work.

Task: pick a real report and plan, in writing, how you would draft, analyse and sanity-check it across three apps. Mark each step where a human must review.

Week 4: agents, Cowork and review (20–25%)

Goal: choose the right tool for a business task and delegate safely.

Where a prebuilt agent helps a workflow, and how to choose between a prebuilt and a custom agent.
Applying an agent from within different Copilot experiences, and filtering chats by agent.
Cowork: multi-step tasks, skills, prebuilt versus custom skills, delegating work, reviewing its plan, progress and results, security considerations and limits, and scheduled tasks.

Task: take the 20-question practice test under a 25-minute limit, mark the domain you missed most and reread that domain guide.

If you are already experienced

If you use Copilot at work every day, two weeks is usually enough: one week on the gaps in fundamentals and responsible AI, and one on the newest material, agents and Cowork, which older tutorials omit. Take the free practice assessment first; if you score well, go straight to the sample questions.

How to know you are ready

You score 80% or more on realistic practice questions twice in a row.
You can explain, without notes, which Copilot tool to use for a given scenario and why.
You can name the check for each risk: inaccuracy, prompt injection, over-reliance, sensitive data.
You finish a 20-question set in about 20 minutes with time to review flagged items.

Use study resources for the official material and the exam tips for the day itself.