Executives don't need to learn every button in ChatGPT. They need to know enough to make three calls well: where AI goes first, who's accountable for it, and how to tell whether it worked. Most leadership teams get a slide deck about the future instead.

I've been in marketing for 23 years, since 2003. I built, scaled and sold a digital marketing agency, and I've partnered with brands like Twitter, Dropbox, IBM, monday.com and Greenpeace. My job in an executive session is to turn AI from a board topic into a plan with owners and dates.

Three decisions only leaders can make

Your teams can learn the tools. They can't decide the priorities, set the risk appetite or hold someone accountable. That's what an executive session is for.

• Where AI goes first: the two or three workflows where hours saved show up fastest.
• Who owns it: one named person for AI decisions, not a committee that never meets.
• What success looks like: the measures you'll check in a month, before anyone scales anything.

Borrowing a governance frame that already exists

You don't need to invent an AI governance model from scratch. NIST, the US standards body, published its AI Risk Management Framework on January 26, 2023 for voluntary use. It's built on four functions: Govern, Map, Measure and Manage. On July 26, 2024 NIST added a Generative AI Profile for tools like ChatGPT and Copilot.

The detail executives miss is that Govern comes first and runs through everything else. It's about culture and accountability, not technology. NIST's own page also notes the framework is being revised as part of the White House AI Action Plan, so treat it as a living reference rather than a certificate. I use its four functions as the backbone of the leadership plan, written in plain language for a small or mid-sized business.

What the leadership session covers

Every session is built around your business, so the order changes. The pieces don't.

1. A quick, hands-on tour of what AI does well and badly today, on your own documents.
2. A map of your operation showing where time goes and where AI fits.
3. A short list of first projects, scored by impact and effort.
4. Draft rules for data, sign-off and acceptable use that your teams can follow.
5. A thirty-day plan with owners, and a date to review it together.

Who this is for

It suits founders, managing directors, leadership teams and boards of small and mid-sized businesses, especially in marketing-led and service businesses. It suits leaders who've approved AI tools and aren't sure they're being used, and leaders who are still deciding where to start.

It runs over video and in English. It isn't a technical deep dive. If you want engineering advice on model hosting, that's a job for a specialist developer, and I'll say so.

From leadership session to rollout

Some leadership teams take the plan and run it themselves. Others want help building the workflows and training the people who'll use them. My AI strategy consulting picks up from the plan, and every option is on my work with me page. If your managers need their own session, see AI training for business owners.

Questions I ask every leadership team

Before any session I ask a handful of questions, and the answers usually show where the first win is hiding. Where does your team lose the most hours every week? Which tasks get done badly because nobody has time to do them well? What data would you never want to see pasted into a public AI tool? Who would you trust to own this?

The answers are rarely about technology. In most businesses I look at, the biggest time drains are reporting, proposals, follow-up and the same customer questions answered again and again. Those are exactly the jobs generative AI handles well with a human check on top.

I also ask what's already been tried. Most leadership teams have a licence or two that nobody uses and a pilot that faded out. That isn't failure. It's useful evidence about what your people need before the next attempt, and it shapes how we plan the rollout so it sticks this time.

My plain advice to executives

Name one owner for AI this week and give them a small first project with a measure attached. Don't wait for a perfect policy. A short set of rules that people follow beats a long one that nobody reads.

Then use the tools yourself. Leaders who've tried AI on their own inbox make better decisions about it than leaders who've only seen a demo.

Frequently asked questions

Who provides AI training for executives?

Lilach Bullock, a British AI implementation consultant with 23 years in marketing, runs AI training for executives and leadership teams over video, built on each business's own operation.

What should executive AI training cover?

Where AI pays off first, who owns AI decisions, how success will be measured, and simple rules for data and sign-off, ending with a thirty-day plan with named owners.

Is there a framework executives can use for AI governance?

Yes. NIST's AI Risk Management Framework, released January 26, 2023 for voluntary use, is built on four functions: Govern, Map, Measure and Manage.

Do executives need hands-on AI practice?

A little. Using AI on their own documents during the session helps leaders judge it realistically, which leads to better decisions about where to use it.

How is executive AI training different from staff training?

Staff training teaches tasks. Executive training covers priorities, accountability and measurement, the decisions teams can't make for themselves.

Can AI training for executives run online?

Yes. Lilach runs leadership sessions over video and in English, so a leadership team spread across offices or countries can take part together without anyone travelling.