Project managers brief people every day: a contractor, an analyst, a new team member. The instruction is clear, the context is given, the expected output is named. Then the work comes back and it gets reviewed. AI assistants need exactly the same thing, and most people give them a fraction of it. The result is generic output, followed by the conclusion that “AI is not very good”.
The BRIEF method fixes that by making the brief explicit. It has five parts and they spell the word.
B — Background
What the assistant needs to know before it can do anything useful. The project, the stakeholders, the constraints, what has already been tried, what changed this week. This is the part people skip most and the one that matters most. An assistant that does not know your sponsor is nervous about the budget will write a cheerful status report.
R — Role
Who the assistant should be while it works. “A senior program manager preparing a steering committee paper” produces different work from “a helpful assistant”. The role sets the standard of judgment and the vocabulary.
I — Intent
What the output is for and what decision it should support. A risk summary written to get a decision from a board is different from one written to brief a new team member. Say which.
E — Expectations
The quality bar and the boundaries. Length, tone, what must be included, what must be left out, what the assistant must never invent. “Do not estimate anything that is not in the data; mark gaps as gaps” is one of the most valuable sentences a project manager can write.
F — Format
The shape of the output. A one-page memo with three headings. A table with these columns. A list of questions. Format is where most of the rework happens when it is left out.
Then: Review
A brief is the start of a loop, not the end of one. The output comes back, it is read critically, and the brief is improved: a missing piece of background, a tighter expectation, a better format. Two rounds of review usually get further than any single perfect prompt. The project manager stays accountable for the result throughout. AI drafts. You decide.
Try it on something real this week: use our free BRIEF Builder to write one brief, run it, review the result, and run it again.
Why project managers are good at this
Because briefing is already their craft. Nobody has to learn a technical skill. They have to notice that the assistant is the fastest, least informed junior they have ever worked with, and brief it the way they would brief a person they respect.
More briefs
Is your project organization ready for AI? Six questions before you buy anything
Readiness is not about tools. It is about data, process, skills, governance, leadership and a first use case worth the trouble.
Governing AI-assisted delivery: what a steering committee should ask
Seven questions a board or steering committee should put to any project that is using AI in its delivery, and the answers that should worry them.
Choosing your first AI use case in a PMO
The best first use case is visible, measurable and slightly boring. Here is how to pick it and how to measure it so the win is unarguable.
What project management interviews ask now
Hiring managers in the US and Canada have changed their questions. Hybrid delivery, AI in the workflow and stakeholder stories dominate. Here is how to prepare.
Tell us about the project, the PMO or the decision in front of you.
A 30-minute call. We listen, we ask questions, and we tell you honestly whether we can help.