Across enterprise learning functions, the consistent pattern is the same: AI tool deployment is ahead of employee capability by a significant margin and the training programmes designed to close that gap are not working.
Not working in a marginal, needs-tweaking way. Working in a ‘we surveyed employees and the majority said the AI training they received doesn’t help them use AI in their actual job’ way. That is not a training problem on the edges. That is a structural failure at the centre of most organisations’ AI strategies.
The gap is not about the technology
It is tempting to frame this as a technology problem. Better tools. More intuitive interfaces. Smarter platforms that surface the right AI capability at the right moment.
But the data and the conversations we have with L&D leaders consistently point somewhere else. The tools are fine. The confidence to use them is not. Most learning leaders know their organisations have adopted AI at pace. Very few feel confident they are building the right capabilities to use it well.
“Insufficient worker skills rank as the top obstacle to integrating AI into existing workflows — not technology limitations, budget constraints, or leadership scepticism.” -Deloitte, State of AI in the Enterprise 2026
That is a learning problem. And it lands squarely in L&D’s domain.
What the readiness gap looks like on the ground
It looks like a workforce that has Microsoft Copilot available in every application they open — and uses it to summarise meeting notes, occasionally, when they remember it is there.
It looks like a compliance team with an AI-assisted drafting tool still writing every policy from scratch, because no one showed them how it works or why they should trust its output.
It looks like an L&D function that has automated content generation — producing more learning faster — without stopping to ask whether the learning it is producing is any more relevant or effective than what it replaced.
Most learning teams are using AI to produce content. Far fewer have fundamentally changed how their function works. More content. Same workflows. Same results.
That is the readiness gap. It does not announce itself. It compounds quietly, every quarter, as the distance between what AI could do and what it actually does gets a little wider.
Why the training that exists is not working
The problem is not that organisations are ignoring AI training. Most are running something. The problem is that what they are running is not changing behaviour.
Generic, not contextual. Most AI training is built around the technology, not the job. Employees learn what AI can do in theory. They do not learn what it can do in their role, in their workflow, with their specific tasks. The result is awareness without application — employees who understand that AI exists and broadly what it is for, but who cannot translate that into Monday morning.
Disconnected, not sustained. A one-hour module followed by nothing is not a learning programme. It is a box tick. Meaningful AI capability requires repeated practice, feedback, and reinforcement in context — the same conditions that build any other complex skill.
Technically focused, not human-centred. The capabilities employees most need alongside AI are not primarily technical. Critical thinking, digital judgement, and the ability to evaluate and challenge AI output are what regulated industries in particular need. These do not come from an awareness module.
The Copilot problem is a perfect example
Microsoft Copilot is probably the most widely deployed AI tool in enterprise right now. Most large UK organisations either have it or are about to get it. And it is, by most accounts, significantly underused.
Not because employees do not want to use it. Not because the technology does not work. But because the gap between ‘here is what Copilot can do’ and ‘here is what it means for you, in your role, tomorrow morning’ has not been bridged.
A generic Copilot training tells a compliance manager that AI can summarise documents. It does not show them how to use it in the specific moments that consume their week: policy review, regulatory mapping, evidence assembly. A finance analyst who knows Copilot can ‘surface insights from data’ is no closer to using it to cut the time they spend on monthly reporting unless someone has worked through exactly what that looks like for them.
That last mile — from capability to application in a specific role — is where most AI learning programmes fail. And it is exactly where the readiness gap lives.
What good looks like
Closing the gap does not require more content. It requires different content, more contextual, more role-specific, more connected to the actual work — delivered in a way that changes behaviour rather than completing a training record.
Role-specific application. The learning objective should not be ‘understand what AI can do.’ It should be ‘use AI to do this specific task better, faster, or more accurately.’ Anchored to a real job task, not a hypothetical one.
Experiential, not passive. People learn to use AI by using AI. The most effective approaches put employees inside realistic scenarios — with feedback, consequences, and the chance to try again. AI tutors and coaching tools are moving in this direction, and the evidence for their effectiveness is growing.
Governance and judgement. Especially in regulated environments, employees need to know not just how to use AI but when to trust it, when to challenge it, and how to take accountability for its output. This is the capability that most generic programmes miss entirely.
What this means for L&D
There is a version of this story where the AI readiness gap is another pressure on an already stretched function. There is a better version.
The AI readiness gap is a problem that sits squarely in L&D’s domain. It cannot be solved by IT or procurement. It requires sustained, contextual, behaviour-change-focused learning capability — exactly what L&D, done properly, exists to provide.
The majority of L&D professionals believe AI will strengthen the function’s influence over the next few years. The ones who will be right are those who move now, diagnosing the gap accurately, designing around real job tasks, and measuring what actually changes in the work.
The readiness gap is not going to close by itself. The organisations that close it intentionally, with a clear strategy and the right learning infrastructure, will be meaningfully ahead of those that wait for the problem to get obvious enough to demand attention.
Where Jam Pan fits
This is the work we do. Not generic AI awareness programmes. Not compliance checkbox training. Not content production for its own sake.
We help organisations diagnose where their AI readiness gap actually sits — by role, by team, by workflow — and design learning that closes it in ways that show up in the work, not just the LMS. Whether that is role-specific capability programmes for regulated workforces, experiential scenario-based learning, or working with your technology platform to surface the right learning at the right moment — this is the territory we operate in.
If you want to understand where your organisation sits and what closing the gap would actually look like, let’s talk.
