Becoming AI-first does not begin with a technology rollout. It begins by changing how people learn, experiment, and improve their work.
Have a conversation with an executive these days and there’s a good chance that the conversation will quickly turn to AI and how it is being used to transform their organisation. Personally, I hear variations of questions like:
- How can we become an AI first organization?
- How can we safeguard our people while building with AI?
- How can we create a culture of AI innovation?
It’s easy to understand why AI is front of mind. Leaders hear about competitors implementing agents, automating workflows, redesigning jobs, and building AI products, and understandably want their organizations to move in the same direction at the same pace. The problem is that all of the competitive noise can make it seem like others are already there, when they simply are not.
McKinsey’s 2025 research found that 71% of organizations regularly use generative AI in at least one business function. That sounds like widespread AI adoption. But there’s a really big difference between employees using ChatGPT internally and meaningfully changing how a business operates. Using a tool is not the same as operating in an AI-enabled business.
So, before you try to transform your entire organization overnight, my recommendation is to slow down and think in baby steps.
You don’t need to do everything at once. What you do need to do is create a few repeatable behaviors that make AI experimentation normal, visible, and connected to the work of the business. Then optimize from there.
If you’re an executive wanting to increase AI adoption across your organization, here are four steps you can start with.
1. Make Your Own AI Learning Visible
If you are telling employees to experiment with AI while never talking about your own experiments, the message could fall flat. Start with simple things. Talk about how you used AI to prepare for a meeting, summarize research, or pressure-test an idea. You are showing your people that learning and experimentation are part of how the work gets done.
For the next two weeks, before you start a task, ask yourself “How can AI help me do this?” Try it, see what happens, then talk about what you learned. You might find it helpful and have a win to share. You might find that it wasted more time than it was worth. Either way, you are modeling the behaviour you want your team to adopt rather than asking everyone else to figure it out.
Once people start seeing you model how to use AI, the next step is giving them the time and space to do the same.
2. Change the Social Signal Around AI
Create a recurring AI showcase, even once a month or once a quarter, where employees share something they have built, tested, automated, or changed with AI. It could be incredibly simple. Someone might show how they reduced a two-hour research process to 30 minutes. Someone else might demonstrate an AI agent they created.
The point is to make learning visible. When people see their coworkers experimenting and sharing, the social risk of trying something new comes down. And that matters because culture is shaped by what people see others doing and whether those behaviours are seen as acceptable and encouraged.
Once you have created a space for ideas to surface, you can start figuring out which ones are worth testing more seriously.
3. Turn Failed Pilots into Training
Once you’ve showcased a variety of ideas, pilot the ones that could actually solve a meaningful problem. You don’t need to worry about each pilot being successful. A failed pilot is actually valuable if you can capture what happened and turn it into a lesson.
Here are some useful things to document after a pilot:
- What did we think would happen?
- What actually happened?
- Where did the workflow fail?
- Was it a technology problem, or did people struggle to understand how to use it?
The answers to these questions can be turned into a case study or training material. And the training can eventually lead to further experiments that actually succeed. Once experimentation, piloting and iterating becomes part of how the organization operates, you need to make sure your leadership team reinforces those behaviours.
4. Put ‘Driving for Change’ as a Leadership Skill in Performance Reviews
If you want to drive AI adoption, you should think about how you will measure that so that senior leaders can be assessed on their performance in driving transformation. You know the saying, if it doesn’t get measured it doesn’t get managed or optimized. Work with your Director of HR to understand how to make this part of the performance review process.
The bottom line
Organizations that are likely to stay competitive are the ones that create an environment where people can learn by experimenting and sharing what they are doing, and ultimately shift how work gets done.
Creating a learning culture starts at the top with senior leaders. To get there, your own learning needs to be visible, your people need to see their coworkers experimenting, the organization should actively learn from failure, and leaders should be assessed on their ability to successfully drive change.
Once you have this, transforming your organization into an AI first organization becomes much easier: one experiment and one changed workflow at a time.
Aneta Kosinska is a Polish – Canadian, MSc + CEMS MIM alumni. With management consulting, analytics and startup/entrepreneurial experience, she aspires to contribute to the knowledge sharing economy.
Image: DALL-E
🔴 Found these ideas useful?
Sharpen your edge in consulting