Programs · Set direction

AI North Star Workshop

What is an AI North Star, and how do you set one? An AI North Star is one business goal so big you can only reach it with AI, written plainly enough that anyone can hold a new AI idea up against it. You set one by putting the people who own the work in a room, drafting candidates, arguing them down to one, and naming who carries it forward.

Moderna's is the one people know: up to 15 new products in five years, when one a decade was normal. A North Star works at any altitude. An enterprise can set one, so can a business area, so can a program team.

What one looks like

A fintech payments platform is aiming for 6x revenue growth on less than double the headcount. You can't staff your way there, which is what makes it a North Star.

Until someone says what AI is for here, every team keeps aiming at its own target.

2 hours · virtual or in person · up to 10 people

What you've built

Your leaders aren't ignoring AI. There's a budget line, there are pilots running, and people who want this to pay off.

The thing most likely holding you back

Nobody has said what AI is for here. Each team picks its own target, so you get random acts of AI. And people fill the silence with worst-case guesses, which is when experimenting stops happening in the open.

The bill for random acts of AI

You pay for the same lesson twice. Two teams build the same thing in the same quarter, and neither one knows.
Ideas get judged by who's asking. With no filter, the sponsor with the most pull usually wins the slot.
People guess. When nobody says what AI is for, plenty of them assume it's for cutting headcount, and the experimenting moves somewhere you can't see it.
The work stays the same. Licenses renew, the usage dashboard looks healthy, and the process underneath hasn't changed.

Who sets an AI North Star

	Enterprise	Business area	Program team
Who's in the room	CEO or COO plus direct reports	Director plus their leaders	RTE, Scrum Masters, Product Managers, Business Owner
Anchors to	company strategy and mission	the enterprise AI North Star, or company strategy if none exists	the area North Star and portfolio strategic themes
The room can commit	workforce, investment and policy	area priorities, area funding, how failed experiments are treated	team norms, capacity for experiments, how learning gets shared

Same two hours at every level. What changes is who's in the room and what that room can promise.

A North Star names the goal one level up. If nothing exists above you, anchor to the strategy that does exist, write down the gap, and send it up. That gap is a finding worth sending.

What I hear from leaders

Questions leaders are asking about setting an AI North Star

We already have OKRs and a strategy. Why do we need one more goal?

Does solving one expensive problem count?

Nobody above us has set one. Can we?

Vision statements, use-case lists and waiting

A vision statement is the common first move. "We will be an AI-first company" is true, and nobody can use it on Tuesday to decide whether a given AI idea is worth the week.

The second is a ranked use-case list. A list sorts what people already thought of. A goal you can't hit without AI changes what gets proposed in the first place.

The third is waiting for the level above you to decide. Most rooms wait a long time. A business area or a program team can set its own, anchor it to the strategy that does exist, and write down the gap above it. That gap is a finding worth sending up.

How many organizations have said what AI is for?

22%

Only 22% of US employees say their organization has communicated a clear plan or strategy for integrating AI. When employees strongly agree leadership has communicated one, they're three times as likely to feel very prepared to work with AI. Source: Gallup, June 2025 (quarterly Gallup Panel survey of US workers)

44%

44% of organizations say AI is scaling across the enterprise. About 6% attribute 5% or more of their EBIT to it, and those organizations are 3.3 times as likely to say they intend to use AI to fundamentally transform the business. Source: McKinsey, The State of AI, August 2026 (1,719 respondents, 97 countries, surveyed May to June 2026)

15 in 5

Moderna set out to bring up to 15 new products to market in five years, and aimed for 100% generative-AI adoption among its digital-enabled people inside six months. Source: OpenAI case study

Once the North Star exists, it becomes a filter question anyone can ask: will this help us get there? Ideas that pass get the time. Ideas that don't get parked or killed in the room.

Start here

Set an AI North Star

Right now

Ask five of your leaders what AI is for here and you get five answers.

With this move

One sentence everyone can check an idea against, with a named owner and a date.

Two hours, up to ten people, one sentence at the end.

The room writes what AI is for here, the philosophy behind it, and a filter question anyone can use on a new AI idea.

How the two hours run

Where are we pointed today?

Everyone writes one sentence on what AI is for here. Read the spread aloud and count the answers. Then both columns go up: what people believe AI is for, and what people worry it's for. Worries stay anonymous.

0:00 to 0:22
Drafting and converging.

Two groups draft candidates, get pushback from Copilot playing your skeptical board member, and revise. The room dot-votes, names the strongest dissent out loud, and merges the wording into one.

0:22 to 1:00
Writing our AI philosophy.

Where do we stand on identity, pace and control? The statement has to answer the top worry from the first round, and it only promises what this room controls.

1:05 to 1:40
Testing the filter and closing.

Run the AI ideas you brought through the filter question. Keep, park, kill. Then rewrite your opening sentence, name the owner, the date, and where this goes next.

1:40 to 2:00

The sponsor sets the altitude with the facilitator and brings the goal one level up.

Five things in your hands at 2:00

Your AI North Star in one sentence
Your AI philosophy: what AI is for, your pace, how you decide, what people can count on
A filter question anyone can use on a new AI idea
Your current AI ideas sorted into keep, park and kill
An owner, a date, and the next audience it goes to

Logistics

$3,500

for up to 10 participants

Format: 2-hour facilitated working session
Delivery: Virtual or in person. In-person delivery adds travel and expenses, billed at cost.
Group: Up to 10 people
Terms: The proposal sets final scope and terms.

Next step: prove it with one team, the Applied AI Workshop

One team takes one real workflow and redesigns it with AI in the room.

Questions about AI North Stars

What is an AI North Star?

One business goal you can only reach with AI, stated so anyone can judge an AI idea against it. Four tests: it's a business goal, it needs AI, you can measure it, and anyone can hold an idea up against it.

How is it different from our vision, mission or OKRs?

Your vision says where the company is headed. Your OKRs say what you'll deliver this quarter. An AI North Star sits above the OKRs and says what your AI work is for, which is the question people are stuck on.

Can a team set one if the company hasn't?

Yes. Anchor it to the strategy you do have, write down what's missing above you, and send that up.

Does solving one expensive problem count?

It can, if it's big enough that you need AI to get there and anyone can tell whether an idea serves it. An executive in one session asked exactly this, and the answer was yes.

Who should be in the room?

The people who can commit at your altitude. At the enterprise that's the CEO or COO and direct reports. In a business area it's the director and their leaders. Up to ten people.