Ninety-four percent of Chief Executives say AI capability is their top in-demand skill. Thirty-five percent believe they’ve actually prepared their people to do it. Eight percent feel very well prepared for the transformation itself. The gap between desire and readiness is now the single largest people risk in financial services, and it’s widening fast.
More than one-third of employees in financial services organisations received zero AI training in 2025. Many people in specialist functions — underwriting, compliance, risk — still don’t understand what artificial intelligence will actually do to their role, let alone how to work alongside it. Meanwhile, the technology keeps accelerating, the board keeps demanding faster deployment, and the people function keeps running a training schedule designed for 2023.
This is not a skills shortage in the market. This is a readiness gap inside your own organisation — between what you’re trying to do, and what your people are actually prepared for.
Why the gap exists
Most organisations have treated AI as a technology programme rather than a people transformation. The Chief Technology Officer has a budget; the Chief People Officer does not. The technology is implemented; training is bolted on afterwards, usually as a one-off. Nobody owns the systemic work of understanding how roles will change, what skills are no longer valuable, which capabilities become essential, and how to move people from the old model to the new one without breaking them in the process.
The default has been: deploy the technology, see what happens, train people to use it. The smarter approach would be: understand the impact on roles and skills, develop people first, then deploy the technology into a ready organisation. Almost nobody does it the second way. It’s too slow, it’s too expensive, it requires the technology and people roadmaps to be in genuine conversation.
There’s also an uncomfortable reality that many boards would rather not think about too hard: artificial intelligence is being deployed to do jobs, which means some of the people doing those jobs, in their current form, may no longer be needed. That’s politically difficult. So instead, the conversation gets fuzzy. We’ll upskill. We’ll redeploy. We’ll find new roles. Maybe. But we won’t be specific, because specificity creates liability.
AI is not about learning to use a new tool. It’s about rethinking what the work is and who does it. Most people functions are still stuck in tool training.
What readiness actually looks like
The organisations that are genuinely ready for AI transformation do the unglamorous work first.
They start with a skills audit. Not “what do we want to be able to do with AI?” but “what can our people actually do today, what is the gap, and what’s realistic to fill?” A compliance officer who has spent thirty years reading regulations is not going to become an AI engineer in a six-week course. But they might become someone who understands what an AI system can and cannot be trusted to do in their function. That’s different, and it’s the actual gap that needs filling.
They redesign roles before they deploy the technology. What does the analyst’s job look like when the tool does the data gathering and the pattern recognition? What does the underwriter do when AI surfaces the risk and they’re responsible for the judgment? These conversations need to happen with the people doing the work, not in a strategy room and then announced afterwards. When people understand in advance how their role is going to change and why, the adoption of the technology is faster and the resistance is lower.
They build AI literacy at every level of management. This is the part most organisations skip or delegate to IT. But managers need to understand enough about AI to have credible conversations with their teams about how the work is changing, what skills are becoming valuable, and what the organisation needs from them. A manager who doesn’t understand AI cannot help their team navigate a transformation that’s fundamentally about it.
63% of employees in organisations with clear role redesign and management preparation adopted new AI tools 63% faster than those where technology was deployed without the supporting people work. The capability gap was the difference between adoption and resistance. — McKinsey, 2025
They create cultural permission to work differently. In risk-averse organisations like financial services, there’s often an unspoken rule: do things the way we’ve always done them. That needs explicit permission to change, and it needs to come from the top. The Chief Executive and the board have to model that it’s okay to try new approaches, to fail sometimes, to learn from the technology rather than just use it the way it’s currently configured.
This is expensive. It’s also far cheaper than deploying AI into an unprepared organisation and discovering six months later that adoption is stuck because people don’t understand why they’re supposed to use it, or don’t trust the outputs, or don’t have any credible sense of how the technology fits into the work they actually do.
The practical starting point
If you’re a CPO in a financial services organisation and your AI transformation is happening faster than your people’s readiness, here’s where to start. (Note: this will feel slow, and everyone will push back on it. Resist. Slowness now saves months of disruption later.)
First, map the impact. For each major function — compliance, risk, underwriting, operations — understand what’s going to change and by how much. Don’t guess. Talk to the technologists and to the people doing the work.
Second, audit readiness. What does that function understand about AI today? What skills are already there? What’s the realistic gap?
Third, design the transition. Not the training — the transition. How do we move from today to the new state without breaking continuity or burning people out?
Fourth, communicate explicitly. People can absorb a lot of change if they understand what’s happening and why. What they can’t absorb is ambiguity about their own role and future.
The organisations that will win on AI are not the ones with the most sophisticated models. They’re the ones with people who understand why the technology matters, how to use it, and whether to trust it. That readiness doesn’t happen accidentally. It happens because someone — usually the CPO, though they need active partnership with the Chief Executive — made the decision that building capability is as important as building tools.
Domi Alzapiedi is a Chief People Officer in banking, focused on the intersection of people strategy, organisational design, and commercial performance. She writes about the questions that keep leadership teams honest.