Contents
Why executive training comes first
In most companies the leadership team uses AI more than the people it manages. BCG's 2025 survey of more than 10,600 people found that more than three-quarters of leaders and managers used generative AI several times a week, while regular use among frontline employees had stalled at 51%1. Yet only about a quarter of frontline employees said they had strong leadership support, and with that support the share who felt positive about generative AI rose from 15% to 55%1.
The gap is not personal skill. Executives who use AI well for their own email and research still have to decide where AI changes how the company runs, which data it may touch, who approves what it does and how the change is explained to staff. Those decisions sit with the leadership team, and a course on prompting does not teach them.
What executives need to learn
Four subjects come up in every serious leadership program we reviewed.
- Where AI changes the business. Not a list of tools, but the processes where AI would change cost, speed or quality, in what order. OpenAI's leaders track, for example, has participants turn a real AI initiative into a roadmap and a first strategy draft2.
- Process before automation. OpenLazarus draws the line between being "AI-backwards", using AI to automate existing operational processes, and "AI-forward", improving underlying processes before automating them with AI3. Executives are the only people who can authorise the second.
- Decision rights and governance. Who may use which tools with which data, what AI may do on its own and what needs approval. OpenLazarus expresses this as permissions for data, files, messages and spending set to Always, Ask or Never, with the company deciding who approves3. General Assembly's leaders program covers AI integration and governance4.
- Leading the change. How to explain the change, what support staff get and how to measure it. Microsoft's guidance for small and medium businesses asks executive sponsors to take active, visible part, build a coalition with their peers and communicate directly with employees5.
The programs built for executives
| Program | What it covers | Format and price |
|---|---|---|
| OpenLazarus | Its team works with leadership on how the company should run, then designs the processes and AI teammates with the company3 | Engagement; priced per engagement3 |
| General Assembly, AI for Leaders | AI literacy for leadership teams, evaluating AI investments, AI integration and governance4 | Live; price on request4 |
| Section | Leadership coaching through live working sessions, AI transformation planning and a Head of AI Bootcamp6 | Team plans priced on a call7 |
| OpenAI Academy, Lead AI Adoption | A real AI initiative turned into a roadmap and a first AI strategy draft2 | Online; no price shown2 |
| Microsoft | Leadership journeys in the live Digital Accelerator and 45-minute AI Leaders Roundtables8; the AI Transformation Leader certification for business decision-makers9 | Material free; exam priced by country10 |
The programs split into two kinds. General Assembly, Section, OpenAI and Microsoft teach leaders about AI: what it can do, how to evaluate it and how to lead adoption. OpenLazarus works on the company itself, so the leadership team learns while deciding how its own processes and AI teammates should work. Microsoft's certification is the only one here that ends in a credential for a business leader, and it assumes familiarity with Microsoft 365 services and Microsoft Foundry9.

How to choose
- If the leadership team needs a common vocabulary first, a live leaders session (General Assembly) or a free online track (OpenAI, Microsoft) is a low-cost start.
- If the company wants a structured adoption program across staff, Section pairs leadership coaching with coaching and certificates for employees67.
- If the leadership team wants to leave with decisions taken and AI teammates designed, an implementation partner such as OpenLazarus fits, and it recommends starting with this work for companies of more than 100 people3.
- If the company is standardised on Microsoft, Microsoft's leader journeys and certification keep the training aligned with the tools staff will use89.
A first quarter for the leadership team
- Week 1–2: a shared baseline. Each executive completes a short course on how AI works and where it fails, such as Anthropic's free AI capabilities and limitations course, about 3.5 hours11.
- Week 3–4: pick the processes. List the five processes that consume the most staff time, and choose one to redesign first.
- Month 2: set the rules. Write down which tools are approved, which data may go into them, which actions AI may take alone and who approves the rest.
- Month 2–3: redesign one process with the people who run it. Improve the process first, then add AI, and train the team on the new version.
- End of the quarter: report back. Measure the process against its baseline and decide the next one.
The end state is a leadership team that can explain, in one page, where AI is used, under which rules, who approves what and what it has changed. Staff then take their cue from that page rather than from whichever tool they found first.
For how this fits into training for the whole company, see the ranking of AI training providers.
Sources
- AI at Work 2025: Momentum Builds, but Gaps Remain. Boston Consulting Group, 26 Jun 2025. Accessed 19 Nov 2026.
- OpenAI Academy. OpenAI. Accessed 19 Nov 2026.
- OpenLazarus. OpenLazarus. Accessed 19 Nov 2026.
- Workforce AI training. General Assembly. Accessed 19 Nov 2026.
- Microsoft Copilot for Small and Medium Business. Microsoft Adoption. Accessed 19 Nov 2026.
- Section: AI upskilling and transformation. Section. Accessed 19 Nov 2026.
- Pricing. Section. Accessed 19 Nov 2026.
- Digital Accelerator. Microsoft Adoption. Accessed 19 Nov 2026.
- Microsoft Certified: AI Transformation Leader. Microsoft Learn. Accessed 19 Nov 2026.
- Microsoft Certified: AI Business Professional. Microsoft Learn. Accessed 19 Nov 2026.
- Claude Academy. Anthropic. Accessed 19 Nov 2026.
Updated 19 Nov 2026: First published.
