Generative AI tools such as ChatGPT, Claude, Gemini and Copilot are useful to project managers for one main reason: a lot of project work is turning messy information into clear structure. Notes into minutes, goals into plans, risks into a register, progress into a status report. AI is good at first drafts of exactly that.
Below are twelve prompts you can adapt. Replace the parts in square brackets with your own details. The more context you give, the better the result.
Before you start: three rules
- Protect confidential data. Follow your company's AI policy. Do not paste client names, personal data or confidential figures into tools your organisation has not approved.
- Treat the output as a draft. AI can sound confident and still be wrong. Check facts, dates and numbers.
- You stay accountable. The AI suggests; the project manager decides.
Planning
1. Turn a goal into a work breakdown
Act as an experienced project manager. Break the following project into phases, deliverables and tasks, in a table with columns Phase, Deliverable, Task and Dependency. Project: [describe the project, goal, team size and deadline].
2. Sanity-check a timeline
Here is my project plan with durations: [paste tasks and durations]. Identify tasks that look underestimated, missing dependencies and anything that could become a bottleneck. Explain your reasoning briefly.
3. Draft a project charter
Write a one-page project charter with sections for purpose, scope, out of scope, key deliverables, milestones, roles and success criteria. Use this information: [paste notes].
Risks
4. Brainstorm risks
List the 10 most likely risks for a project like this: [short description]. For each, give the likely cause, impact (high/medium/low) and one mitigation action. Format as a table.
5. Write a risk escalation
Draft a short, calm email to the project sponsor explaining this risk: [describe]. Include the impact on scope, time or budget, the options we have, and my recommendation. Keep it under 200 words.
Communication and reporting
6. Status report from bullet points
Turn these notes into a weekly status report with sections: overall status (green/amber/red), progress this week, next week, risks and issues, decisions needed. Notes: [paste notes].
7. Adapt one message for different audiences
Rewrite this update in two versions: one for executives (three bullet points, focus on impact and decisions) and one for the delivery team (practical next steps). Update: [paste text].
8. Difficult stakeholder email
A stakeholder has asked for [new request] which is out of scope. Draft a polite, firm reply that acknowledges the request, explains the impact, and proposes a change request process.
Meetings
9. Agenda with time boxes
Create a 45-minute agenda for a [type of meeting] with these goals: [goals]. Include time boxes, owners and the decision we need at the end.
10. Minutes and action items
From these rough meeting notes, write concise minutes: key decisions, open questions, and an action list with owner and due date. Notes: [paste notes].
Thinking partner
11. Lessons learned
We just finished this project: [summary of what went well and badly]. Help me write a lessons-learned summary with concrete recommendations for the next project.
12. Challenge your plan
Act as a sceptical steering committee member. Ask me the ten toughest questions about this plan: [paste plan summary].
Getting better results
- Give a role and a format. "Act as …" and "Format as a table" improve almost every answer.
- Iterate. Reply with "shorter", "more formal", or "add a column for owner" instead of starting again.
- Save prompts that work. Build a personal library for recurring tasks such as weekly reports.
For a complete approach to using AI across the project life cycle, from planning to automation and team adoption, see AI for Project Managers.
