The board is asking about AI and you don’t have a view

The question comes up at every board meeting now. What are we doing about AI?

Most chief executives answer it the same way: the team’s been told to try things. Some of them have. A few emails got quicker. Someone built something clever that nobody else uses. A training day happened and nothing much changed afterwards.

That isn’t a strategy. It’s a hope that someone below you will work it out.

Why “go and play” isn’t a strategy

Trying things is fine. Trying things with no question to answer isn’t.

• Nobody knows what good looks like. So nobody can tell whether it’s working, and the experiments never turn into decisions.
• The keen people race ahead and the nervous ones dig in. Some of your team will quietly put up barriers, because they’re worried about what it means for their job. Telling them to play doesn’t deal with that.
• It shows up in front of customers first. The AI-written email that sounds like nobody in particular. The proposal that reads like everyone else’s. Buyers can tell.
• You still don’t have an answer for the board. Just a list of things people have tried.

Where AI helps a B2B firm first

Usually not where people start.

Most businesses start with content, because it’s the easiest thing to try. It’s also where AI helps you least. AI is brilliant at the obvious, so ask it to write about your business and you’ll get the expected version of your category.

In a B2B firm of 50 to 200 people, the useful places tend to be less visible:

• The report nobody has time to build. Somebody in your business spends a day pulling numbers together so you can see what happened last month. That’s often the first thing worth connecting.
• Systems that don’t talk to each other. The CRM, the website, finance, and the spreadsheets that sit between them. Joining them up is what lets you answer “is this working?” without waiting a week.
• The jobs your team does by hand every week. Follow-ups, handovers, the same email written forty times.
• How AI tools describe you to buyers. Your buyers are already asking ChatGPT who to shortlist. What it says about you is part of your reputation now. More on that in who’s writing your brand story if you’re not.

Each of those makes the business better, not just faster. They’re also far easier to measure than a pile of AI-written blog posts.

What has to be true first

AI gives you more of whatever you’ve already got. If what you’ve got is muddled, you get more muddle.

Two things decide whether AI helps or just adds noise.

Your data. If the numbers live in five places and nobody trusts any of them, AI won’t fix that. It’ll give you confident answers built on bad information. Connecting the data comes before automating anything.

Your positioning. If your leadership team describes the business three different ways, AI will produce all three, at scale. It writes well from a clear position. It can’t decide the position for you. If you’re not sure yours is clear, these four tests will tell you.

That’s why the first step is rarely a tool. It’s agreeing what the business is for, and getting the data into one place you can trust.

A 90-day plan you can present

The board doesn’t need a five-year AI strategy. It needs to see you’re in control, with a first step and a way to measure it.

Days 1 to 30: find out where you are. Ask your team, privately, where they’re already using AI and what worries them. Map which systems hold which data, and where someone rebuilds a report by hand. Ask ChatGPT and Perplexity what they say about you and your nearest competitors, and save the answers.

Days 31 to 60: pick one or two places. Choose where AI would make the business better and where you can measure the change. One internal (a report, a manual process) and one external (how you show up to buyers) is a good balance. Agree what success looks like before anyone starts.

Days 61 to 90: do it properly and report back. Make the change with the people who’ll use it, not around them. Then take the board one page: what you tried, what changed against the baseline, and what you’ll do next.

That’s a plan you can stand behind, because it’s built on your business rather than someone else’s framework. And if the honest answer to one of the board’s questions is “we don’t know yet”, say so. Nobody has all the answers on AI. The leaders who look in control are the ones with a way of finding out.