We are at an inflection point for AI. The question is no longer whether your organization is adopting it. It’s whether your people are actually capable of using it. Most aren’t.
This isn’t a technology failure. The tools work. The problem is simpler, yet harder to fix now. Companies deployed AI before they built the people capable of using it.
At Docebo, we help enterprises build workforces that can actually use AI. We surveyed 2,000 people to find out where adoption breaks down, and the bottleneck shows up in an unexpected place.
The challenge with AI adoption isn’t one problem. It’s a compounding series of them, each one making the next harder to solve. The numbers bear that out: 56% of workers are so buried in manual, pre-AI tasks that they have no time to learn the tools designed to free them from those tasks. That’s the first wall.
The second is when they do find time, 85% can’t connect what they learned to their actual role.
Finally, 78% say training happens in systems completely disconnected from where they actually work.
Three walls. Each one built by the same decision: moving fast on tools and slow on people. The result is a workforce that has access to AI and no real ability to use it. You can’t hand someone a tool they don’t know how to use and call it transformation.
Organizations moved quickly on tools, announcements and initiatives. They did not move quickly on building the human capacity to use them. The infrastructure employees rely on to upskill wasn’t built for this rate of change. It was built for a slower, more predictable era of skill development. That era is over.
And yet most companies are still running the same playbook: Deploy the tool. Schedule the training. Check the box. Move on.
That approach doesn’t produce capability. It produces completion rates. Those are not the same thing. Most organizations are measuring the wrong thing. Seats purchased. Licenses deployed. Modules completed. These are procurement metrics dressed up as readiness metrics. They tell you what your organization spent. They say nothing about what your people can actually do. The real measurement is evidence. Did the person actually demonstrate the capability in their work, on the systems where work happens, in a way that holds up to a regulator or a CFO?
Real readiness is a skills question. Can this person apply what they learned to their specific work, their specific goals, right now? If you can’t answer that, you don’t have an AI strategy. You have an AI expense.
The fix isn’t better content. It isn’t more training hours. It’s a fundamentally different philosophy for how learning works inside an organization. The organizations getting returns aren’t the ones with the most licenses or the most content.
The successful organizations are connecting learning to an individual’s specific role and goals. They’re embedding learning AI into the culture of how work gets done.
That shift requires deliberate choices.
Build pathways that move with the employee
AI literacy isn’t one skill. It’s dozens, and they surface differently by role. A sales rep using AI for proposal research hits different gaps than a manager interpreting AI-generated performance insights or an analyst pressure-testing AI outputs before acting on them. Map learning to those moments. When someone uses a new AI feature for the first time, that’s the window, not six weeks later in a scheduled training. Tie progression to real work and something changes. People stay when they can see where they’re going.
Make every expert part of the infrastructure
In most organizations, a small number of people have already figured out how to use AI tools effectively. They’ve developed workarounds and instincts no formal curriculum has captured. Find them. Assign them as mentors during tool rollouts. Build short peer-led sessions around how they actually work. The goal isn’t a new program. It’s turning knowledge that already exists into something transferable.
Embed learning inside the tools employees already use
Don’t ask people to leave their workflow to build an AI skill. Surface it where the work happens: a prompt in your CRM when a rep uses an AI-generated summary for the first time, a triggered module when someone accesses a new feature, a nudge when a team adopts an AI-assisted process. Friction is where learning dies. Remove it.
Give employees real ownership with a clear starting point
Ownership without direction isn’t empowerment. It’s abandonment. Start with a skills assessment tied to how AI is changing their work specifically, not generic digital literacy, but the capabilities the role requires now and will require in 18 months. Show people the gap. That visibility is what turns self-directed learning from an aspiration into a habit.
Capabilities like these require infrastructure most organizations are still building. A persistent learner profile that follows the employee from role to role. Skills intelligence that infers capability continuously, not on an annual review cycle. Learning that lives inside the tools where work actually happens. Most companies have none of these. The ones that build it will.
The organizations ahead of this understood early on that you can’t play catch-up on readiness after the fact. People’s ability to use AI tools has to develop alongside the tools themselves. Every day that gap widens, it gets harder to close.
But capability doesn’t compound on its own. It compounds when it sits on top of data no AI agent can fabricate. Compliance records tied to specific people on specific dates.
Skills graphs that reflect what your workforce can actually do. Learning history across years and roles. External training data tied to revenue, certification, and customer outcomes.
That’s the foundation that makes readiness durable. Without it, you’re running training cycles faster and measuring the wrong thing.
The companies that lead through this shift won’t be the ones that moved fastest on AI tools. They’ll be the ones that built the data foundation and the human capability to use it together. That’s the only advantage that compounds.
