The pattern seemed inevitable: new technology arrives, requires technical expertise, and creates a divide between specialists and everyone else. The people with deep domain knowledge, i.e. the policy analysts who understood legislative nuance, the engineers who knew where inefficiencies hid, the procurement officers who could sense risk in supplier bids, remained separated from the systems that could amplify their expertise because they needed translators, IT tickets, and lengthy approval processes just to access the tools that could help them work more effectively.

AI was supposed to follow the same pattern, with early predictions warning of a new digital divide between AI-fluent workers and everyone else. But something unexpected is unfolding: GenAI can actually do the opposite, closing gaps rather than creating them, and the reason is fundamental to how these new tools work.

When technology uses human language rather than requiring specialised syntax, the barriers that separated technical from non-technical workers for decades collapse. You don't learn programming languages or master complex interfaces, you simply describe what you need the way you'd explain it to a colleague, and the system executes it. The policy analyst can now query complex databases conversationally. The manufacturing engineer can build predictive models by describing patterns they've observed over years. The procurement lead can automate complex evaluations by articulating their decision criteria in natural language.

Yes, effective prompting technique still makes a significant difference in the quality of outputs you get from AI, but learning good prompting is fundamentally easier than learning SQL or Python. It's a communication skill, not a technical one, which means the barrier to competence has dropped dramatically for people who were previously excluded from working directly with data and systems.

This represents genuine democratisation, because the people who understand business problems most deeply can finally build solutions directly, without translation layers or technical intermediaries.

The leapfrogging opportunity most organisations are missing

There's a powerful parallel here to what happened with mobile telephony in developing economies. Countries across Africa and Asia leapfrogged landline infrastructure entirely, moving directly from no phones to mobile networks because building mobile infrastructure was both cheaper and more effective than installing copper wire to every home. They didn't have to go through the intermediate step that developed nations took.

AI creates exactly the same leapfrogging opportunity for workforce capability. Your employees don't need to follow the traditional path of Excel, then SQL, then Python, then finally AI. They can skip the SQL and Python entirely and go straight to conversational AI interfaces that let them work with data directly, leapfrogging straight to the capability that actually matters: getting insights from data and systems using natural language.

But here's what we see consistently when organisations try to capture this opportunity: they approach AI adoption the same way they approached every previous technology wave. They identify their most digitally fluent employees, train them extensively, designate them as AI champions, and hope knowledge spreads organically to everyone else.

In pursuing this familiar strategy, organisations recreate exactly the divide that AI's conversational interface has the potential to eliminate. McKinsey's November 2025 State of AI report revealed the scale of this problem: while 88% of organisations now use AI in at least one business function, only 39% report any measurable impact on their bottom line.

When you focus AI adoption on super-users and centres of excellence, you're investing in the 15% who were already digitally fluent rather than unlocking the 85% who weren't. The same people who navigated previous technical systems now navigate AI systems with confidence. Everyone else watches from the sidelines, convinced once again that they're "not technical enough" despite the fact that technical skills are no longer the barrier.

Building an AI-native workforce means including your whole workforce

From our work helping organisations move from AI tool deployment to measurable adoption, one principle matters more than any other: to build an AI-native workforce, you need to include your whole workforce from the start, not just small champion teams who will supposedly cascade knowledge later.

This isn't just about fairness or inclusion, though those matter of course. It's about unlocking the expertise that already exists throughout your organisation but has been trapped because people lacked the technical skills to apply or scale it directly. Your subject matter experts (the people who deeply understand your customers, your processes, your risks, your operations) are the ones who should be using AI most intensively, because they're the ones who know which questions to ask and how to interpret the answers in business context.

A subject matter expert with deep domain knowledge plus a generative AI licence is an incredibly powerful combination, often underestimated. But it doesn't happen automatically just because you've deployed the tools. You need to invest deliberately in upskilling people so they develop confidence and capability to use these tools effectively in their actual work.

This is particularly important because there's significant fear around AI, especially among people who haven't historically considered themselves digitally literate. They've watched technology waves pass them by before, and might worry AI will make them irrelevant. This fear prevents them from engaging with the technology that could actually make them more valuable than ever.

The opportunity here is profound: AI can help people who felt left behind by previous digital transformations feel relevant again. When someone who struggled with Excel discovers they can analyse data conversationally, when someone who never learned to code can build workflow automations by describing what they need, it's transformative. They're applying the expertise they already have more effectively, and that changes their relationship with technology fundamentally.

But capturing this requires deliberate effort. You can't just give people licences and hope they figure it out. You need structured support that builds confidence through successful application to real work problems. You need to show them explicitly that this technology is for them, not just for the technical people they've always relied on.

The strategic choice ahead

AI genuinely can reverse the digital divide that widened over the past years, offering organisations a rare opportunity to unlock expertise that has been trapped throughout their workforce for decades. The barriers separating domain experts from the systems that could amplify their impact have collapsed, and the leapfrogging opportunity to skip entire generations of technical training is real and immediate.

But that potential isn't automatic. Organisations that focus on super-users and centres of excellence are recreating the exact divide AI makes unnecessary, solving for concentrated expertise when the genuine opportunity lies in distributed capability across everyone who understands your business deeply.

The fundamental divide is no longer between those who can code and those who can't. It's between organisations that recognise AI as a strategic opportunity to empower their subject matter experts directly and those that treat it as another technology to be mastered by specialists before eventually trickling down to everyone else.

A year from now, that strategic choice will be unmistakable in productivity metrics, employee engagement/satisfaction scores, and competitive market positioning. The organisations that built widespread capability, that trusted and empowered their subject matter experts with proper support and structured upskilling, that helped people who felt left behind by previous digital transformations discover they can thrive with AI, will be pulling away from competitors still running pilot programmes.

The question isn't whether your people are ready for AI, because they already possess the most important capability: deep understanding of your business and the ability to communicate clearly about what they need. The question is whether your organisation is ready to unlock that potential by including your whole workforce in this transformation, not just the digitally fluent minority who've always been comfortable with new technology. It’s time to re-imagine how work gets done. Together.