The people side of bunq's AI transformation

What it takes for people to make AI work at scale, with Bianca Zwart, Chief Strategy Officer at bunq

Jul 02, 2026

👋 Hey, Dan here.

Last week, in my conversation with Bianca Zwart, I shared how bunq has redesigned its entire structure and how it operates around AI.

That piece was about the systems. This one is about the people.

We talk a lot about the tech side of AI adoption, but we don’t talk enough about the human side.

It’s one thing to optimise a process with AI. It’s another thing to get your team to unlearn how they’ve always worked, adopt a new mindset, and be excited about the change.

(If you’re a founder or operator struggling with this right now, you are not alone)

I wanted to explore how the best teams are tackling this shift.

Read it if you want to know:

Whether AI is replacing people at bunq
Why AI is not owned by one team at bunq
How non-engineers won bunq’s internal AI hackathon
What bunq looks for in hiring now
What Bianca personally had to relearn about how she works

Bianca Zwart is Chief Strategy Officer at bunq, the Dutch neobank founded in 2012 that now serves 22 million users around the world.

She also leads bunq’s largest division – operations – overseeing more than 300 people.

She describes her role as making sure bunq stays user-centric. And AI, in her view, is what makes that even more possible at scale.

“The question we ask at bunq is not humans or AI. It’s what will solve this user’s problem properly and get them to the right outcome faster?”

AI-first is not easy for people

I asked Bianca directly: Has the transition to AI-first been easy for people at bunq?

She didn’t pretend it was. And was clear about what the hardest part has been:

“The most difficult part is getting experts to redesign their process in a way that’s actually AI-first.”

She’s had to make the shift herself. Which she says was not easy.

Even as the person architecting the shift across the entire company, she still had to personally unlearn and relearn how she works.

She points to a concrete example:

“I was originally looking at things through a process lens rather than an outcome lens. For example in Ops, I realised we were just using AI to get through a checklist faster, instead of optimising for the actual goal for our users.”

“So I had to shift from ‘How can we make this faster?’ But ‘Why do we have this in the first place?’ And ‘Should we?’”

What has made it easier for Bianca, is that for her, it connects back to why bunq started in the first place.

“If I think back to when we started bunq, the whole reason we started was because we thought things could be done better. We always had an aversion to ‘things have always been done this way.’”

”AI is really good at forcing you to confront your own assumptions. So people are challenged to think that way again.”

Adoption follows evidence

Shifting to AI-first isn’t a switch you flip once, that stays flipped. It’s something people have to keep doing, day after day, until the new way of working becomes the default.

“I think with any habit, it’s a discipline in the way you work. You need to remind yourself constantly that it’s different, that you can do it differently.”

And people won’t truly adopt a new habit just because they’re told to. They adopt it because they feel what it’s like to work the old way, and now the new way, and notice the difference.

What makes people stick with it, in Bianca’s experience, is seeing the payoff for themselves.

“Once people see the outcome of working that way: how much better their work becomes, or how much faster they can reach their goal, once they see the impact, they’re very much inclined to want to keep working that way.”

“I think that’s the most interesting change on the human side that I’m seeing, and everybody’s going through it. The great thing about people at bunq, is that they’re very intrinsically motivated to change and to learn. So it’s been more exciting rather than challenging.”

Build ownership around “why”

bunq doesn’t organise around job titles. Instead, everyone has an “ownership.”

“An ownership is the defined problem for our users that you’re expected to solve. We write an ownership document for every ownership, and the first question to answer is: ‘Why?’ Then: ‘What is the problem you’re solving for our users?’”

“With AI, we can better optimise for that outcome. And the why of your ownership becomes the core of the AI agent you build to help execute the outcome.”

Here’s a short extract Bianca shared from someone at bunq with user feedback ownership:

Why: To make my life easy, you need to make my bunq experience frictionless.
Whenever there’s friction, you need to remove it. Show me that you care to give me the best experience, and that you swiftly adapt to meet my needs. Closely listen to my feedback and act upon it.
What: When something in my bunq experience breaks, confuses, or annoys me, I expect you to:
Notice it the very first time it happens.
Fix the root-cause fast, so my peers and I never bump into it again.
You act as the second line of defence: when any ownership within bunq fails to address my struggles, you step in and ensure that each and every one of them has a timely, verified, structural fix.

Why AI is not owned by one team at bunq

bunq used to have a dedicated AI team. Not anymore. Hans Scheffer also got rid of HelloPrint’s dedicated AI team.

“We used to have that, roughly a year ago. But that’s no longer the case. Now AI is embedded in every team. Everyone is AI-led.”

“We do have a thin AI engineering team, but their role is to enable the rest of the company. They build the platform, the tools, the standards that let every team work the same way. They’re not a separate AI island doing AI on everyone else’s behalf. It’s not owned by one team. It’s part of how every team operates.”

“AI handles the vast majority of the work, and whatever needs to be routed to an expert gets routed to an expert.”

What does real adoption look like? Like non-engineers winning hackathons

A clear proof point of widespread AI adoption came at the most recent “bunq Weekend” – a company retreat where bunq’s entire team flies in from all over the world.

One of the planned activities was a 24-hour AI hackathon.

“What struck me most is that at the end of the AI-athon, most people on stage presenting their solutions were not engineers. People from operations, for example, built working AI agents themselves, without a single engineer involved. It was amazing to see.”

The standout example came from an entirely non-engineering team who built an AI agent that lets anyone at bunq instantly ask what users think: about a feature, a direction, a pain point – and get a synthesised answer drawn from eight different feedback sources at once.

As Bianca put it:

“User insight used to be fragmented: scattered across interviews, surveys, and support signals that meant manually cross-referencing. With this agent, we can sense-check any decision against the real friction points users face across the different stages of their lives - whether that’s moving abroad, starting a business, or anything else.”

I asked Bianca whether it was still being used after the hackathon.

“For sure it’s still being used. And if anything, it will be used more and more across the company, as it provides us with a continuous gut-check that keeps product decisions and all other teams grounded in what users actually want and need, not assumptions.”

The shift shows up clearest in Support

I asked Bianca which department has been most transformed by AI:

“AI has completely changed how our Support teams work. A lot of time used to go into repetitive tasks like searching for information, drafting replies, or manually processing requests.”

“Today, AI handles most of that heavy lifting, so our people can focus on solving the most complex cases.”

AI agents now handle a lot of the high-volume, repetitive work: such as refunding invoices or filing bug reports.

Before, a support team member would spend time gathering information piece by piece just to file a bug report. Now an AI agent handles that whole conversation and produces a complete, structured report.

The human job is to review it – and with the time freed up, they focus on what matters more: analyzing interactions, identifying improvement points, making sure users leave a conversation happier than when they arrived.

“When AI can produce an answer in seconds, the human’s role is no longer just to have the answer. It is to understand the user well enough to know whether it is the right one, to apply judgment and to spot what the machine has missed. And, most importantly, to ask better questions.”

Is AI replacing people at bunq?

I asked Bianca if AI is replacing people’s jobs at bunq?

“Actually, it’s the opposite: despite our efficiencies, our staff rate is stable, and we’re still hiring in line with the company expansion.”

“The things that make someone great at their job: product intuition, creativity, judgment, empathy – those aren’t things AI can replicate.”

“The whole ‘AI is replacing jobs’ conversation tends to happen when companies treat AI as the strategy. The goal becomes making the company more efficient. But that’s not how we think about it.”

“Our users are the strategy. AI is just the best tool we’ve ever had to serve them better and stay human as we scale.”

“The question we ask isn’t ‘what can AI do instead of people?’ – it’s ‘what can our people do now that they couldn’t before?’”

What bunq looks for when hiring now

I asked Bianca whether who they hire has changed – if the bar has been raised because of AI.

“It’s not so much that the hiring bar has changed – we’ve always hired people who love solving problems, not to fill roles. What’s changed is the tools.”

“AI can take care of much of the execution now, so we’re looking for people who are curious, adaptable and take ownership. The people who thrive at bunq today aren’t just great at getting things done; they’re great at asking the right questions, challenging assumptions, and using AI to find better solutions, faster.”

“We care less about what someone already knows and more about how quickly they learn, adapt, and turn new technology into a better experience for our users.”

I asked what the most important skills they look for now are:

“The most important skills for applicants right now are three things: critical thinking, resilience, and no ego.”

“AI is making knowledge more accessible and more of the execution can be automated. What can’t be automated is someone who spots a flawed assumption in a data model, changes course when a product bet does not land, or brings human empathy to a complex decision.”

“What matters more now is whether you can question an assumption, let go of your preferred solution when something better comes along. Ultimately, it is about the problem at hand, not who is or is not solving it. The people who thrive care more about finding the best solution, more than being the person who had the idea.”

Bianca’s favourite interview question tests exactly that:

“I end every interview the same way. I ask candidates: ‘Based on everything that we’ve discussed, everything you’ve read, everything you heard, can you give me one reason why you would not be a good fit for this role?’”

She says the answer usually says more than a CV ever could.

“You would be surprised that the more senior people get, the harder it gets to answer this question.”

AI hasn’t replaced expertise, it’s amplified it

I asked Bianca whether one function has changed more than others because of AI.

“It’s not so much about how one role changes more than another. AI hasn’t replaced expertise, it’s amplified it.”

“The value people bring today is increasingly in judgment, creativity, and asking the right questions. It’s less about how roles have changed and more about how we all use the tools to find better ways of solving the problems at the core of our ownerships.”

The clearest shift she does point to is in management itself.

“As AI helps people execute faster, managers become less focused on coordination and more focused on creating direction, developing talent, and enabling their teams.”

Nor does she expect people to keep pace by mastering every new release.

“We don’t expect people to know every new tool or model – that’s impossible. What matters is staying curious, experimenting, and continuously improving.”

Bianca believes the people who thrive in an AI-dominated world will not necessarily be the ones who know the most.

“They’ll be the ones who understand how their role is changing, and use AI to become more user-centric, more effective, and better at solving the right problems.”

Thanks for reading In Founders Words! This post is public so feel free to share it.

AI is moving fast. You’re adapting. Your role is shifting. In all that momentum, it’s easy to forget to ask: where am I actually going with this?

That’s the question I’ve been sitting with in my coaching sessions with Rose Marie van Leeuwen.

I’ve been working with her for six months now, and what she’s good at is helping you slow down to see clearly – so that when you speed up, you’re heading somewhere that actually makes sense for you.

She works with founders and professionals on inner clarity: cutting through noise, getting out of your own way, and figuring out what you actually want from all this change.

If you’re in the middle of a big shift right now – professionally or personally – Rose Marie is worth a conversation.

Thanks, Bianca! You can also read Part 1 of this series, about how AI has completely changed bunq, as well as Bianca’s Un-Dutch Playbook: How To Build with Speed (like bunq)

Have a great week, and keep on building.

Until next time,
Dan 👋

In Founders Words

Discussion about this post

Ready for more?