Many companies approach AI training the same way they do compliance modules. Schedule a course. Check a box. Move on. And then wonder why nothing changes.
What I have learned is that the only time you can use AI to its full potential is when you are changing how people perform at work. Training is just a tool, not the goal. The real difference isn’t “How do we train people on AI?” It’s “How do we allow people to optimize their ability to use AI?”
That reframe makes a big difference. People will not change their behavior through content alone. A person must be motivated, know how, and be in an environment that supports the development of the desired behaviors. Therefore, pushing courses on people should be the last thing you do and not the first. Learning & Development and Human Resources departments have a much larger role than creating curricula. They need to provide an environment that makes working with AI the easiest, most intuitive way to do great work and achieve better outcomes, including value types that were not possible before.
Standardize the Foundation So Teams Can Safely Move Fast
Companies that race to apply AI successfully identify a clear approach: centralized consistency where needed and decentralized flexibility where appropriate.
Centralize and standardize consistency across all employees, including leadership buy-in and expectations, tool access and provisioning, rules governing data use and privacy, and general principles of responsible AI use. This forms the foundation for successful AI initiatives. If this foundation does not exist, you get chaos.
Decentralize and create flexibility for teams to experiment with AI and test different use cases within each team’s workflow, and provide an environment for local sharing of learning experiences.
Think about this process in terms of infrastructure. The company builds the interstate system so people can travel safely at high speeds. Individual teams design the local roads that fit their specific context. HR and L&D are key partners in building that interstate and making sure every team can actually drive on it.
Not Everyone Needs to Be an Expert on Day One
AI maturity is not a binary switch that flips from ‘untrained’ to ‘expert.’ In reality, people progress along a continuum. Using clear stages can make AI feel more approachable by giving employees a path they can see themselves on.
We use three stages to think about this:
- First is capability: people are experimenting with AI, building their confidence, and leveraging it for specific tasks.
- Second is adoptive: People begin to streamline their workflows by creating reusable workflows rather than using AI one-off.
- Third is the redesign stage: People begin to rethink their entire workflows and systems to create new value at scale.
Fluency has multiple components, too. It’s not just about knowing which buttons to press. To be proficient, people need to develop the appropriate mindset and strategic thinking, practical skills, and the ability to collaborate effectively with AI. Also, people need to consider the ethical implications of using AI. Developing all these aspects takes time and practice.
Your role is not to push every person to proficiency. Your role is to provide a clear path forward for individuals and support them at each stage of development. Sharing these stages and components openly also helps employees self-assess where they are today and choose their next step, rather than feeling that AI proficiency is a vague or hidden target.
Skip the Lecture Hall
Don’t start by putting everyone through the same formalized training. Instead, think of how you can help people develop experience using AI to do real work in their real jobs.
Establish an AI champions network. Every organization has people who are already experimenting on their own. Find out who they are and give them space to demo what they’ve done, tell stories about what worked (and what didn’t), and share templates with their colleagues.
Provide a safe space to experiment. Similar to hack weeks or AI labs, where people get 8-10 hours to apply AI to their own workflows. Emphasize that this is about the learning process and sharing information, not demonstrating immediate returns on investment.
Curate libraries of templates, use cases, and resources so people don’t drown in information. A small library of prompts, checklists, and workflows tied to real roles is worth more than a hundred generic articles. Help people feel caught up instead of overwhelmed.
Develop a plan for employees to learn from each other. Peer-to-peer learning works best when there are regular opportunities for employees to share knowledge and experience. Show-and-Tell style meetings, or “Lightning Talks,” work well. Just keep the format simple. Share how you used to do things before you started using AI. Share how you’re doing things now with AI. Finally, share what you learned along the way. That’s it.
Success Is About Changed Work, Not Completed Courses
Success should be measured by results (outcomes), rather than by how many people complete the training or by the number of hours of content viewed.
Instead of focusing on the numbers, focus on the outcomes. For example, how many business processes were changed and/or improved through the use of AI? What percentage of the company’s departments use at least one AI-supported process that has measurably improved speed, quality, or value creation? How often are employees actually using the tools provided, and what level of confidence do employees have when completing a task that requires the utilization of those tools?
Collecting concrete examples will also provide evidence. Examples include time savings on specific tasks, improved quality of deliverables, and new value created that could not have been produced before AI implementation. This is where the true story lies.
This is not about checking a training box; it is about enabling your employees to perform better. By keeping your eye on this goal, the rest typically follows.
Colin Monaghan is the Director of Learning & Development at Zapier, where he focuses on helping people do their best work in one of the world’s most recognized fully remote companies. He lives in the Pacific Northwest, where he balances screen time with trail time, a deep reading habit, and an unapologetic baseball obsession.
