Your AI Leadership Plan

One page, not a strategy deck

You now have the pieces: a working model of what AI does well and badly, six questions for any pitch, a way to choose where to invest, a governance approach, and a message for your people. This lesson turns them into one page you can put in front of a board, a leadership team or your whole organisation.

One page is deliberate. A long AI strategy gets praised and shelved. A single page gets read, challenged and acted on. It forces you to make choices rather than list possibilities, and it makes it obvious when something important is missing.

What the page must contain

Why, in terms of problems. Two or three sentences on the business problems you are addressing. Not "we need an AI strategy", but the specific costs, delays or pressures you want to move.

Two or three pilots. Each with a process, a named owner, a measure agreed up front, today's baseline and a review date. More than three and none of them gets proper attention.

Governance decisions. Who owns outputs, which data may go into which tools, what has gone to legal for a view, and where the AI register lives.

People commitments. What you have told staff, what you will not use AI for, how training will work, and how people will hear about changes that affect them.

What you are not doing. A short list of things explicitly out of scope. This prevents assumptions and protects the pilots from scope creep.

Review. When the leadership team will look at the whole plan again, and what would make you change course.

The template

AI LEADERSHIP PLAN
Organisation: [name]     Owner: [you]     Date: [today]


WHY (the problems, not the technology)
[Two or three sentences. Which costs, delays or pressures are we addressing?]


PILOTS (no more than three)
Pilot A
Process:        [what, who does it today, how often]
Owner:          [one named person]
Measure:        [the number that should change]
Baseline today: [measured before we start]
Human check:    [who reviews output, at what point]
Review date:    [6 to 10 weeks out]  Decision: scale / change / stop

Pilot B
[same fields]


GOVERNANCE
Output ownership: every output has a named human owner.
Data rules:       [which data may go into which approved tools]
Legal view:       [questions sent to legal or compliance, and status]
AI register:      [where it lives, who maintains it]


PEOPLE
What we have told staff: [one line]
We will not use AI for:  [e.g. hiring, pay, performance, redundancy]
Training:                [how, on what, with protected time]
Changes affecting roles: [how and when people will hear]


NOT DOING
[Two or three things explicitly out of scope]


REVIEW
Leadership review of this plan: [date]
We would change course if:     [one or two conditions]

Pressure-test it before the board does

Every board or leadership team will ask a version of the same questions. What is this costing, what happens when it goes wrong, how will we know it worked, and who is accountable? Have a plain answer to each and the conversation becomes a decision rather than a debate.

Here is my one-page AI leadership plan:
[paste it]
Act as a sceptical but supportive non-executive director.
Ask me the six hardest questions about this plan, hardest first.
For each, describe a weak answer and what a strong answer must include.
Then list anything in the plan that is a decoration rather than a decision.
Do not rewrite the plan.

Presenting it

Lead with the problems, not the technology. Directors engage with a delay, a cost or a risk they already recognise. Then walk through the pilots, and say plainly that stopping any of them is a real outcome. That one sentence tends to make the rest of the plan more credible, because it shows you are running experiments rather than defending a commitment.

Be open about what you do not yet know. If legal has not answered a question, say so and give the date you expect a view. A board trusts a plan with visible gaps and dates for closing them far more than one that appears to have every answer.

Finally, ask for a decision, not a discussion. Approve the pilots, agree the review date, and name who reports back.

Where to go next

Your managers need practical skills for running AI day to day in their teams: choosing tasks, estimating value, reviewing output. Point them to AI for Managers.

For the rules that keep adoption safe at team level, including what staff can share and a no blame reporting line, use Safe AI for Teams.

If agents and automated processes are on your roadmap, the Agentic AI hub covers how they work and where they go wrong.

And if you want structured training for a whole team or leadership group, see our partners page.

Pass the quiz below and the Clear Strategy badge is yours. More usefully, so is a plan you can take into your next leadership meeting.