We’re not prepared for the hardest part of the AI transition. It’s not the technology. It’s the people. So far in this series, I’ve written about Agentic AI in financial services — how it will redefine operating leverage, and the governance architecture required to make it work inside regulated institutions. But this is the question every leadership team is quietly wrestling with: what happens to the humans? I’ve saved this for now not because it’s secondary — but because it’s the most consequential question of all. And it deserves more than a reassuring paragraph buried inside a discussion about efficiency ratios.

I lived through one of the most significant technology transformations in banking — at BBVA. When we started, fewer than one in ten product sales were digital. Within a few years it was nearly six in ten, and eventually almost ninety percent of retail banking activity moved to the app. Digital customers grew to nearly thirty million. NPS reached number one across eight markets. Customer attrition dropped meaningfully. Those numbers are real. I was inside them. And what I can tell you is this: the technology was never the hard part. The hard part was the organisation — the culture, the identity, the people who had built careers doing things one way and were now being asked to do something fundamentally different. Every metric we celebrated externally had a human story behind it: teams unlearning as much as they learned, managers leading through uncertainty they couldn’t fully resolve, colleagues who needed to be brought forward rather than left behind. Agentic AI is that same dynamic. But this time the shift is larger, the speed is faster, and the stakes are higher. What got us here won’t get us there.

Here’s what I hear consistently from CEOs and senior leaders: technology readiness is further ahead than most admit publicly, and human readiness is further behind than most admit privately. That gap is where most transformations will fail. Because there are two very different versions of this transformation. The technology version: deploy models, build infrastructure, report productivity gains. The organisational version: redesign how work happens, redefine roles, rebuild how people create value. Most institutions think they’re doing the second. In reality, they’re still operating in the first. Most AI strategies today are technology roadmaps disguised as transformation plans. That’s the mistake. I’ve been in conversations with CEOs who are genuinely wrestling with this — not just strategically, but personally. They know this will change roles their organisations have had for decades. And they know that how they manage that change will define not just performance, but their institutional identity.

What we’re actually witnessing is the emergence of a new form of capital: Agent Capital. It won’t replace Human Capital — but it will redefine it. At its best, Agent Capital isn’t about substitution. It’s about amplification. The institutions that win won’t be the ones with the most AI. They’ll be the ones that figure out how to build Agent Capital without breaking Human Capital. Think about how banks have always measured the value of their people — workforce cost set against return on equity, headcount against efficiency ratio. The same discipline now needs to apply to a new class of participant. What is the return on each agent deployed? How do you measure agent productivity, agent utilisation, the efficiency ratio of a hybrid team? This is what we call Return on Agent Capital — ROAC — and it will become as fundamental to institutional performance management as ROE is today. The leaders who start building that measurement capability now will have a structural advantage over those who treat agent deployment simply as a technology cost line.

This is where Paloma Tejada’s work is critical. She has spent years at the intersection of organisational design and transformation, and her point is simple: “You can’t deploy agentic systems into a static organisation. The moment you introduce them, the organisation itself has to change.” — Paloma Tejada Gasset. This is the part most roadmaps miss. Institutions plan the deployment. They don’t plan the organisation that has to receive it. But when agents take on structured execution, what changes isn’t just cost — what changes is the nature of work itself. And that requires a fundamentally different approach to transformation. Not technical. Human.

The technology changes. The people don’t. In every transformation, three types of people show up. There are those who lean in immediately — the leapers. They experiment before being asked, they see possibility rather than threat, and they are your early proof that change is possible. Then there are the bridge builders. The people who are up for innovation but they are cautious. Bridge builders will do deep research and once convinced they bring others with them. And there are those who hold onto the current model — the tradition holders. Often deeply experienced, always resistant, but not always wrong. In many cases, they are protecting something genuinely important. Most organisations over-index on the leapers and under-invest in the bridge builders and tradition holders. The organisations that succeed build bridges, not just momentum — because the failure mode in most transformations isn’t the technology. It’s the culture fracture.

There is a natural instinct in most institutions to sequence this kind of change: get the technology working first, then once it’s stable, turn to the people side. It feels logical. It is also where things start to break. The organisations that navigate this well approach it differently. They design the agentic workflows and the human development paths in parallel — not as separate workstreams but as a single, integrated question: what does the new model require of our people, and what do our people need in order to thrive within it? That discipline mattered at BBVA. It matters even more now, because the window to get it right is much shorter.

The most important roles emerging right now aren’t just technical. They’re human — the people who sit at the boundary between agent output and human judgment, who know when to trust the system, when to challenge it, and when to override it. As Paloma puts it: “What drives roles now is the data and who makes the final decision. That shifts the entire conversation about accountability — and about what it means to be a manager when part of your team isn’t human.” — Paloma Tejada Gasset. This is new leadership territory. Managing hybrid teams — humans and agents — requires systems thinking, judgment, and a level of accountability most management roles haven’t historically required. These are not simpler roles. They are harder ones. And they will define regulatory trust, customer outcomes, and institutional integrity.

Every technology wave has displaced work. Every wave has also created new forms of work. What changes is the nature of human contribution. In the agentic era, that contribution becomes more centred on judgment, relationships, and accountability. The repetitive coordination overhead, the documentation burden, the low-value orchestration that consumes so much time today — those begin to move to machines. What remains is the part of the work that actually requires humans. And in the best implementations, organisations are honest with their people about what is coming, and make visible commitments to bringing them forward — equipping them for a future that is coming — rather than simply deploying around them. Their judgment carries further. Their capacity expands. Their impact compounds. The future of banking won’t be defined by AI alone. It will be defined by how institutions learn to combine Agent Capital and Human Capital — building organisations where technology amplifies people, and where Agent Capital elevates Human Capital rather than diminishing it. That’s the real transformation. And it’s exactly the problem we’re building to solve.