AI Executive Workshop: Recap of My Recent Session
This week I ran an AI executive workshop for senior leaders from different organisations across Germany. One room. A mix of sectors and roles. Everyone there to answer the same question: what do we actually do about AI?
This was not a coding class. It was a working session for decision-makers. We covered mindset, the current state of AI, real use cases, and hands-on tasks the leaders ran themselves. Below is a recap of what we did, why it worked, and what each person took home.
What an AI executive workshop is
An AI executive workshop is a leadership session, not technical training. It is built for CEOs, board members, department heads, and functional managers.
The focus sits on five things:
- Mindset: how leaders think about AI and change.
- Strategy: how AI supports business goals.
- Real workflows: how to use AI on everyday tasks.
- Governance and risk: how to adopt AI safely.
- Decisions: how to choose pilots, budgets, and partners.
The aim is simple. Help leaders talk about AI in business terms, tell hype from reality, and make faster, calmer decisions.
Who was in the room
The group came from several organisations across Germany. Different sectors. Different job titles. A common gap.
Most had already tried AI in some form. Few had a shared plan. That is the pattern I see almost everywhere now. The market has moved past the question of “is AI real” and into the harder question of “how do we run this well.”
People in the room included founders and executive directors, plus heads of HR, finance, sales, operations, and IT. The people who set strategy and hold the budget.
The problem most leaders walk in with
Many teams are busy with AI but short on clarity. They lack a clear answer on four points:
- Strategy: how does AI fit our business goals?
- Governance: who owns AI decisions, policy, and risk?
- Investment: where do we spend time and money first?
- Impact: what real value will this deliver?
Mindset adds to the muddle. Some leaders see AI as a threat. Some treat it as a magic fix. Many quietly file it under “someone else’s job.” That mix slows everything down.
So we started there.
What we covered
I keep the structure tight so leaders leave with a plan, not just notes. Here is the ground we covered.
Mindset and change
We opened with how leaders think about AI. Fear, hype, doubt, and opportunity all showed up in the room, which is normal.
The leaders shared their own views and how they want to guide their teams through change. The shift I look for is from “AI is a threat” to “AI is a tool we can shape.” That shift sets the tone for everything after it.
The current AI landscape
Next, a plain overview of where the technology stands. No jargon for its own sake.
We looked at the main types of AI and where each fits in a business:
- Generative AI: writing, summaries, research, chat.
- Predictive AI: forecasting, risk scoring, churn.
- Automation: document processing and routine tasks.
- AI agents: tools that plan and act across several steps.
The point is not to know how each one works under the hood. It is to know which one suits which job.
Strategy and use cases
We then tied AI to the business. Leaders mapped real problems in their own organisations and tested each idea against three questions:
- What is the business impact on revenue, cost, time, or quality?
- Is it feasible with the data, skills, and tools we have?
- What is the risk to privacy, security, and reputation?
From that, each leader built a short list of high-value use cases to start with. A small, sharp list beats a long wish list every time.
Risk and governance
We covered the parts that keep boards awake. Data privacy. Security. Bias. Wrong or made-up answers from a model. Compliance under EU rules.
Then we set the start of a governance approach: who owns what, basic rules for data and model use, and how to handle issues when they come up. This is what turns AI from a side project into something a board can stand behind.
Four ideas that reframed the room
Some sessions turn on a handful of ideas that change how leaders see the whole field. Four did the heavy lifting here.
Agents: AI that does, not just answers
I touched on agents earlier. They deserve more than a line.
Most people meet AI as a chat box. You ask, it answers. An agent goes further. You give it a goal, and it plans the steps and acts to reach it.
Think of the gap between a calculator and an assistant. A calculator waits for the next sum. An assistant takes the job off your desk and comes back when it is done.
Why it matters: this is where real process change comes from. Agents can move work through several steps, not just draft one paragraph. For leaders, that turns AI from a personal helper into a way to run parts of a process.
Tokens: how AI reads, writes, and bills
Every model works in tokens. A token is a small piece of text, often part of a word. The model reads tokens in and writes tokens out. OpenAI provides a useful explanation of how tokens work and why they affect both cost and context limits.
Two reasons leaders should care:
- Cost. You pay by the token. More text in and more text out means a bigger bill.
- Memory. A model can only hold so many tokens at once. That is its working memory for a task. Push past it and it starts to lose the early detail.
Why it matters: once leaders grasp tokens, AI cost and AI limits stop being a mystery. They can budget and plan instead of guess.
A harness: the model is the engine, not the car
A model on its own is just an engine. Powerful, but it cannot get you anywhere alone.
The harness is everything you build around it. The tools it can use. The data it can see. The rules it must follow. The memory it keeps. The checks on its work.
Picture a Formula 1 engine sitting on the floor. Impressive, and useless without the car around it.
Why it matters: the winners will not be the firms with the best model. Everyone can buy the same model. The edge comes from the harness you build around it, using your own data and your own processes.
A new learning for the whole organisation
The last idea decides whether any of this sticks. AI skill cannot live with one team. The whole organisation has to learn, and keep learning.
We have seen this before. When email and the internet arrived, everyone had to adjust, not just IT.
Why it matters: adoption stalls when only a few people are fluent. When the whole organisation learns together, AI becomes part of how the place works, not a side project for the keen few.
The hands-on part, where it got real
This is the section leaders remember. We stopped talking about AI and used it.
Each person worked on their own real tasks, not made-up examples. They saw the time saving in front of them, in minutes. Tasks included:
- Reporting: summarising long reports and pulling key points from documents.
- Research: building quick briefs on competitors and market trends.
- Analysis: comparing options and stress-testing scenarios, such as “what if costs rise 10 percent?”
- Presentations: drafting outlines and slide content.
- Planning: sketching roadmaps and resource plans.
- Customer comms: drafting emails and proposals, then tailoring the tone for different readers.
The reaction was the same one I see in every good session. Quiet at first, then a lift in the room as people realise the work they dread can take a fraction of the time, at a higher standard.
What the leaders left with
A workshop is only as good as what people carry out the door. This group left with:
- A shared, clear view of what AI can do for their organisation.
- A calmer, more practical mindset toward AI.
- A short list of high-value use cases tied to strategy.
- A first governance approach and a view of the main risks.
- A draft adoption roadmap with phases: pilot, scale, embed.
- A 30 to 90 day action plan with owners and dates.
- Real, practical skills from using AI on their own tasks.
No one left with a pile of theory. They left with a plan they could start on Monday.
Why I run workshops this way as an AI keynote speaker in 2026
I run two businesses. A staffing and AI firm, and a not-for-profit AI education organisation, AI Ireland. That means I sit on both sides of the table every week. I see what works in real projects, what fails, and why.
That is the gap I try to close in the room. Plenty of AI content is hype or pure theory. Leaders do not need more of either. They need someone who has reviewed real projects, trained thousands of leaders, and can translate AI into business decisions.
Buyers have noticed the shift too. The demand now is for sessions that move a team from experimenting to operating. Less show. More plan.
How to choose an AI executive workshop
If you are weighing up a session for your own leadership team, use this short checklist.
- Business-focused, not theoretical. It should work on real problems and end with actions, not just awareness.
- Mindset plus hands-on. Leaders should use AI on real tasks during the session, not just watch a demo.
- Tailored to you. It should reflect your sector, size, data, and rules. Avoid one-size-fits-all content.
- Real examples. Ask for cases close to your industry and region.
- Ends with a plan. Look for a clear roadmap and a 30 to 90 day action plan with named owners.
A good session changes how a team decides, not just what it knows.
FAQ
What is an AI executive workshop?
A leadership session that helps senior decision-makers understand AI in business terms, find high-value use cases, and agree a practical adoption plan. It is strategic, not a coding class.
Who should attend?
CEOs, founders, board members, and heads of HR, finance, sales, marketing, operations, and IT. Anyone who sets strategy or holds budget.
How long does it run?
Formats vary. A 3 hour session gets a leadership team up to speed fast. A 6 hour session can produce a full adoption roadmap. Longer programmes run over weeks with follow-up support.
Is it technical or strategic?
Strategic. The goal is better decisions, a clearer mindset, and alignment at the top, with enough hands-on work to build real confidence.
What do people leave with?
A shared view of AI opportunities, a shortlist of use cases, a first governance approach, an adoption roadmap, a 30 to 90 day action plan, and practical skills from hands-on work.
A final word
The German session followed the same arc I see across Ireland and Europe. Leaders walk in unsure and a little wary. They leave with a clear list, a plan, and the confidence to act.
That is the whole point. Move a leadership team from curiosity to action in a single day.
If you want to run a session like this with your own leadership team, get in touch through markkellyai.com.
About the author
Mark Kelly is an AI keynote speaker and board advisor. He has delivered more than 300 keynotes for organisations including Microsoft, Oracle, Salesforce, and PwC, trained over 10,000 business leaders, and reviewed more than 1,000 real AI projects through the AI Awards programme. He is the founder of AI Ireland and co-founder of an AI and cybersecurity staffing firm, which gives him a working view of both AI strategy and delivery.
