AI for Project Managers: Status Reports and Risk Registers

Every project manager I have ever taught has the same Thursday afternoon. The work is going fine. The team knows what it is doing. But somebody upstairs wants a status report by Friday morning, and now you are scrolling back through three weeks of meeting notes, a Teams thread, and a task list that has not been updated since the 12th, trying to reconstruct what actually happened. That reconstruction job is where AI for project managers earns its keep — not in some grand replacement of your judgment, but in the two hours a week you spend retyping things you already know.

I want to be specific here, because most advice on this topic is useless. “Use AI to save time” is not advice. So here are two things you can do this afternoon, with the exact wording, and what goes wrong if you skip a step.

Tip 1: AI for Project Managers Starts With the Status Report

This works in Microsoft Copilot, ChatGPT, or Claude — the trick is not the tool, it is the prompt. The mistake people make is pasting their notes and typing “write a status report.” You get back something that reads like a press release and says nothing. You have to hand it the format.

Gather your raw material first: standup notes, the meeting transcript Teams generated for you, and your task list. Copy them into one block of text. Then use a prompt shaped like this:

You are helping me write a weekly project status report. Below are my raw notes from the past three weeks. Produce a report with exactly these sections: (1) Summary, three sentences maximum. (2) Completed This Period, as a bulleted list. (3) In Progress, as a table with columns Task, Owner, Target Date. (4) Blockers and Decisions Needed. (5) Risks. Use only information that appears in my notes. If a date or an owner is not stated, write UNKNOWN — do not estimate or fill it in.

Then paste the notes underneath. What comes back is a draft with the structure already built, the owners pulled out of the text, and — this is the part people underestimate — a list of everything you thought you knew but never actually wrote down, flagged as UNKNOWN. That list is genuinely useful. It is your call sheet for the next twenty minutes.

The gotcha: that last sentence of the prompt is not optional. Without it, every one of these tools will quietly invent a target date or assign a task to whoever was mentioned nearby, and it will look completely plausible. I have watched a class send a report with a fabricated milestone date because it read naturally and nobody questioned it. Tell it to mark gaps, then check the gaps yourself.

Tip 2: Build a real risk register from the task list you already have

Most risk registers are written once at kickoff and never touched again, because updating one by hand is miserable. This is the second place AI for project managers pays off, and it takes about ten minutes.

Start with the task list you already maintain. In Microsoft Project, File > Export, or just select the Task Name, Start, Finish, Resource Names, and Predecessors columns and copy. In Planner, use Export plan to Excel. In Excel, select the range and copy. You want plain text or CSV — not a screenshot.

Paste it in with this:

Here is my project task list with dependencies and owners. Identify the ten items most likely to cause a schedule slip. For each one, give me: the risk in one sentence, why this specific task is exposed (single owner, long dependency chain, external dependency, no slack, or unclear scope), a likelihood of High/Medium/Low, and one concrete mitigation I could put in place this week. Return it as a table. Rank by likelihood.

What you get back is a systematic read of your own dependency structure — the pass you would have done yourself if you had a quiet hour. It catches the things that are obvious in hindsight: the one person who is the only named resource across six tasks, the handoff that depends on an outside vendor, the four tasks stacked back to back with no float.

The gotcha: strip anything confidential before you paste. Real client names, contract values, staff performance comments. Replace them with “Client A” and “Vendor B.” The analysis works exactly as well on placeholders, and you have not put anything into a chat window that your compliance officer would like to discuss with you. And treat the output as a first draft — you still own the judgment about which risks are real.

Plenty more on the calendar

If you want to see this done live, with the prompts on screen and real project files rather than tidy examples, I am running How to Use ChatGPT & Microsoft Copilot For Project Management on Wednesday, October 7. We cover status reporting, risk registers, meeting summaries, and the places where these tools get it wrong.

The full current schedule is on the front page of PCWebinars.com, with new dates going up as they are confirmed. Register for this session, or browse the others — there is usually something on Excel, Power BI, Copilot, or Claude within the next few weeks.

About your trainer

I’m Tom Fragale. I have been a computer trainer and consultant for more than 30 years, and I have taught well over 30,000 people how to get more out of the software they already own — Excel, Access, Power BI, Word, Outlook, and now the AI tools that sit on top of all of it: Claude, ChatGPT, and Microsoft Copilot. My background is applied, not theoretical. Alongside teaching, I build real databases and reporting systems for clients in Excel, Access, and Power BI, so the examples I use in class are the problems I ran into last week, not textbook exercises. I also publish free tutorials on YouTube and longer courses online.

You can find the full schedule of upcoming live sessions, along with recordings and course listings, at PCWebinars.com.