Your team already uses AI, but nobody talks about It: How one AI Practice Day changes that and get your team on the same page regarding AI usage.

In almost every project team I work with, AI is already part of daily work. Yet hardly anyone knows who uses it, for which tasks, and with what results. One colleague asks ChatGPT to summarise meeting notes. Another drafts status emails with Copilot. A third generates weekly status reports. And some team members have not touched AI at all, because they don’t know where to start. Each person develops their own habits, collects their own experience and gets different results. AI has arrived in the team, but as a collection of silent individual experiments.

Many leaders respond by starting with policies, tool decisions or a company-wide AI training. From my experience: general AI training rarely changes habits. You need to connect the „training“ with workflows and tasks from daily work to see the benefits and see how AI works. A better first step is to look at how the team works today. Which tasks take time? Where do routines repeat? Where can AI help, and where does it only look attractive?

In this article, I explain how a single AI Practice Day can align a team on AI usage, create a shared direction and deliver first solutions people.

 

 

The problem: AI use without a shared direction

 

When people use AI on their own, without shared rules and often without the project lead knowing, we speak of shadow AI. It rarely comes from bad intentions. People want to get their work done, and freely available tools make that easy.

The consequences are less harmless than they look.

• Data risk: confidential project, client or personal data can end up in tools that were never approved.
• Uneven quality: two people doing the same task with AI produce very different results, and nobody knows which approach works better.
• Lost knowledge: good practice stays with individuals. What one person has figured out never reaches the rest of the team.
• Divided teams: some colleagues move ahead, others hold back, and the gap between them grows.

 

The root cause is rarely the technology. In most teams I see three things coming together.

First, AI arrived bottom-up. People started using tools long before anyone defined how the team should work with them. By the time leaders pay attention, habits are already in place.

Second, nobody creates room to talk about it. People don’t share their AI use because they are unsure whether it is allowed, or because they worry it makes their work look less valuable. So experience stays hidden.

Third, organisations try to solve the problem from the top. They write guidelines, select tools or launch training programmes, but these rarely connect to the specific tasks a project team handles every week. People attend the training and return to their old habits.

Banning AI does not fix this, and neither does looking away. What helps is making existing AI use visible and developing it further as a team, starting with the work people do every day.

 

 

My approach: the AI Practice Day

 

An AI Practice Day is a single workshop day with a team. It does not require a large AI initiative, a technical background or a finished strategy. It requires a team, its everyday tasks and a clear structure for the day.

The day follows two phases.

 

Phase 1: Find where AI helps

The morning focuses on the team’s tasks. Everyone brings the recurring work that takes up their week: status reports, meeting minutes, stakeholder emails, collecting risks, preparing reports. The team looks at these tasks together and identifies where AI can reduce effort, improve quality or catch mistakes.

This is also the moment when hidden AI use comes to the surface. Someone mentions they have been using Copilot for a task for months, and a colleague asks how that works. These moments are valuable, because they show what already works in your context. For this to happen, people need to feel safe. Nobody should fear consequences for sharing how they work today.

The team then prioritises. Some ideas bring high value and are easy to implement. Others matter strategically but need more preparation, data access or budget. Some are simply not worth the effort. By the end of the morning, the team has agreed on a short list of three to five use cases it wants to pursue, together with a clear view of where a person stays responsible for the result.

A side effect is just as important: the team develops a shared view of what AI can and cannot do. Sceptics see where it helps. Enthusiasts see where it falls short.

 

Phase 2: Build first working solutions

The afternoon turns the priorities into something usable. The team learns a shared way of instructing AI, so results become comparable and good prompts can be passed from one person to another. The colleagues who are already more experienced play an important role here, because the team learns from them directly.

From there, the team builds a first reusable assistant for one recurring task. This sounds technical, but in most cases it requires no programming.

 

 

What the day delivers

 

By the end of the day, the team has:

• A shared picture of where AI helps in its daily work, and where it doesn’t
• A short, agreed list of use cases the team stands behind
• A common way of working with AI, so knowledge can be shared
• A first prototype of a reusable assistant, tested with the team’s own material
• Clear next steps: what to scale, what to explore further and what to park for now

The most important result is alignment. Before the day, everyone worked alone, with their own tools and methods. Afterwards, the team works from the same starting point and can learn from each other. A team that shares what it learns improves week by week. Individuals working alone start from scratch every time.

 

 

Common mistakes to avoid

 

In my client work, I see the same mistakes again and again, and most of them happen before the workshop even starts.

Many teams begin with the question of which AI tool to use. These discussions rarely lead anywhere, because the team has not yet agreed on which problems it wants to solve. Once the tasks and priorities are clear, the tool question usually becomes much easier to answer.

Another common mistake is inviting only the colleagues who are already enthusiastic about AI. The day feels productive, but the results often stay within this small group. The sceptics are the ones who later decide whether a solution is used in daily work, so their voice belongs in the room from the beginning. Their questions also make the results more robust.

Teams also tend to aim too high. With so many ideas on the table, it is tempting to start several solutions at once. In practice, one assistant that works reliably builds far more trust than several half-finished ones, and it gives the team a clear success to build on.

I also see many workshops end without a clear decision on what happens next. The energy in the room is high, people go back to their desks, and within a few weeks the results fade. A short decision at the end of the day about what to scale, what to explore further and what to park makes the difference between a good workshop and lasting change.

Finally, some leaders expect too much from a single day. A practice day creates direction, shared understanding and first prototypes. Integration into internal systems, work with sensitive data and binding governance rules need further steps, and they need IT and data protection at the table.

 

 

What corporate leaders can do now

 

• Make existing AI use visible in your teams, and create a safe environment where people can talk openly about how they already use it.
• Start with daily work, asking which recurring tasks cost your teams the most time before you discuss tools or policies.
• Run one AI Practice Day with a pilot team and use what you learn to decide how to roll it out further.
• Protect the time, so the day does not become extra work on top of full calendars.
• Keep people accountable for every AI-supported result, and make clear where review and approval stay with the team.
• Turn individual knowledge into team standards, so good practice spreads instead of staying with a few people.
• Treat the day as a starting point, and involve IT, data protection and other functions once prototypes move towards productive use.

 

If you want to identify and prioritise the right AI use cases for your teams with a clear structure, the AI Practice Day is an effortless start. Get in contact to learn more.