Your team already uses AI. That's not the same as enablement.
- Daily AI use and effective AI use are two different problems. The first is adoption. The second is enablement.
- When asked how they wanted to learn, employees ranked a live workshop last.
- What they wanted was to learn from each other: shared prompts, real examples, and a look at how colleagues work.
- Share the results back with the team and decide together. For us, it led to an updated AI policy.
I surveyed our team on how people are actually using AI at work. Over half the team responded, and 82% of them said they use it every day.
The workflows ranged widely. Some people had built an agent that reads their inbox and summarizes their day: what came in, what they've answered, what's still waiting. Others had added a brand guideline plugin, so every document they create starts out in the right branding instead of getting fixed at the end.
The number that stuck with me came next. Almost everyone said they're probably leaving value on the table.
Adoption isn't the problem anymore
The point of the survey was to find out where we stood. The answer was clear. We're past adoption. People use AI daily without being told to. What they don't have is a way to get better at it.
That's enablement, and it's a different problem with a different fix. Most companies are still treating AI as an adoption problem: approving tools and encouraging people to try them. If your team is already there, more encouragement won't help.
They didn't ask for a workshop
When I asked what kind of training would help most, a live workshop came in last.
What ranked highest:
- A shared library of the prompts people are already using
- Real examples from other teams
- A chance to see how colleagues actually work with AI
Nobody asked to be taught AI. They asked to learn from each other.
The tools are changing quickly, but so is the way people want to learn them.
That makes sense. The tools change so fast that a training deck is out of date by the time it's delivered. A prompt a colleague used yesterday on the same kind of work is specific, current, and already tested.
What we did with it
I presented the results to the team, and we talked through what would actually work. Then we updated our AI policy to give everyone clearer guidance on how to use AI internally.
That discussion mattered as much as the data. People could see the gap themselves, and the next step came from the team instead of being handed to them.
How to run this at your company
- Ask how often people use AI, and for what. Collect the actual workflows, not a yes or no.
- Ask whether they think they're getting full value. That's where the gap shows up.
- Ask how they want to learn, and let them rank the options. Live workshop, prompt library, team examples, office hours. Don't assume.
- Share the results and decide together. Then update your policy and enablement plan to match how people actually work.
What this means for People leaders
AI enablement is landing on People teams whether we planned for it or not. It's a learning and development problem, and the tools are changing faster than any curriculum can keep up. Ask first, build around how people already work, and expect to revisit it often.
It's the same lesson I wrote about in Hire someone or buy a tool? Usually neither. Before you build something new, find out what's already working.
Building AI enablement into how your team works?
View company services