Open a few project risk registers and you will often see the same lines. “Resources may be unavailable.” “Scope may change.” “Delays may occur.”
These are not wrong. They are just not useful. Nobody can act on them.
I’m not a project manager. I build AI prompt systems, so for this guide I studied how risk work is done across a whole project, from the first brainstorm to the closing review, and where it tends to go weak. Three weak points came up most:
- Risks are too generic to act on.
- Scores come from gut feeling, with no clear scale behind them.
- Reports are too long for the people who need to decide.
The first one has a known cause. Risk identification suffers from group think, time pressure and blind spots. When the kickoff meeting is short, the team writes the obvious risks and stops.
Where AI Helps, and Where It Doesn’t
AI can widen your view. It can list risks you did not think about, challenge your optimism, and write a first draft of a risk document in minutes.
It cannot know your sponsor’s appetite for risk, your team’s real capacity, or what is in your contract beyond the text you give it. It can also make arithmetic mistakes, so check any number it calculates, especially in cost and schedule risk.
What did not work: one-line questions like “What are the risks of my project?” Generic question, generic list. Once I added the project type, team size, budget and deadline, the list became much more specific.
One Prompt to Try Today
This is Prompt 3 from the book:
“Act as a devil’s advocate. I am about to present this project plan to leadership and claim it is low risk. Challenge my assumptions and identify the five most dangerous hidden risks I may be underestimating. [Paste project plan summary]”
It works because it asks for a clear number (five) and gives the AI a job: argue against you. The book suggests using it before executive presentations, to test for optimism bias.
Compare “Delays may occur” with the example output for Prompt 1 in the book: “Key subject matter experts are shared with another project and may not be available during critical testing phases, potentially delaying go-live by two to four weeks.” One names the cause and the size of the impact. You can plan for the second one.
👉 [Get AI Prompts for Project Risk Management on Amazon]
Before and After
Weak prompt: “List the risks for my project.”
Better prompt (my example, based on Prompt 1): “I am managing a software implementation with 8 team members, a budget of [amount] and a deadline of [date]. List risks across technical, schedule, budget, resource, stakeholder, compliance and external categories.”
The second one gives the AI something to work with.
How the Book Is Organised
AI Prompts for Project Risk Management has 142 prompts in 15 chapters:
- Chapters 1 to 3: identifying, categorising and scoring risks
- Chapter 4: quantitative analysis, including expected monetary value, schedule risk, Monte Carlo concepts and contingency reserves
- Chapters 5 to 7: mitigation planning, risk communication, monitoring and tracking
- Chapters 8 to 10: stakeholders and escalation, risk reviews and lessons learned, risk culture and PMO governance
- Chapters 11 and 12: Agile and hybrid risk, plus advanced strategies such as scenario planning and portfolio risk
- Chapters 13 to 15: specialised scenarios (vendor, regulatory, cybersecurity, people), ready-made communication templates, and a reference library
There are also 3 bonus sections: 6 quick-start templates, a 5-step guide to adapt the prompts to your organisation’s terms and scoring scale, and a phase-by-phase checklist.
Who It’s For
Project managers, program managers, PMO staff, Agile leads, and business analysts who keep a risk log. Also team leads who were asked to “look after the risks” without much guidance.
Use It Carefully
Review every output with your team before you present it as final. Some prompts ask you to paste a contract or vendor proposal. Check your company policy first, or remove names and figures. And I cannot promise that better prompts will protect your project. Risk work still depends on people making decisions.
Start With the Phase You Are In
You do not need to read in order. Bonus Section 3 lists which prompts to run in each of the 4 project phases. Phase 1 has 9 prompts. The book suggests picking one live project and running that Phase 1 checklist this week.
👉 [Grab your copy of AI Prompts for Project Risk Management]
If you want the basic prompt formula behind all my guides, it is in a free book: AI Superpower for Everyone.
