Here’s how human-centered AI will transform financial services
When I left the boardroom of ING, one of Europe’s most powerful banks, for a small, half-empty office in a forgotten corner of Amsterdam, I had nothing but two leather chairs, a vision and a handful of algorithms. Many thought I had lost my mind.
In truth, I had found my purpose. When I led the daily banking domain at ING and shaped strategies for the next generation, I kept running into an uncomfortable truth: we’re graduating brilliant young people into a financial system they are not prepared to navigate. Not because they lack talent or ambition, but because we don’t teach them the necessary skills – and because access and advice remain out of reach for far too many.
The problem is global. Billions of people lack the knowledge or tools to make informed financial decisions. ‘Thin-file’ customers – those about whom there is little financial data, including students, recent immigrants, gig workers and sole traders – number about 2 billion worldwide.
There are also problems around small businesses, the engines of jobs and economic growth. Sixty-five million of the world’s 360 million formal SMEs face an access-to-finance gap, driving an annual shortfall of $5.2 trillion (or $8 trillion when informal enterprises are counted). This isn’t about uncreditworthy customers; it’s about outdated models that can’t see their potential.
That realization collided with another. Inside large, listed institutions, short-term numbers can eclipse long-term value. Innovation stalls, customers go underserved, trust erodes. I wasn’t frustrated with any one person or company; I was frustrated with a system that too often fails the people it’s meant to serve.
So from that small Amsterdam office, we built tools to address those systemic failings. We started with eyeOpen, one of the first robo-mortgage advisers; then we launched AdviceRobo, using psychometrics and behavioral signals to rethink risk and inclusion.
The lesson that changed everything was this: beneath every data point is a person. When finance learns to see the person, everything else follows. Data alone doesn’t make lives better – understanding does.
This conviction is what I call Banking 5.0, a strategic shift from legacy, product-centric models to a human-centered system, where AI plus behavioral economics delivers services that are both smarter and more empathetic.
The emergence of Banking 5.0
Banking has always evolved with technology. Paper-based ledgers gave way to mainframes and ATMs (Banking 2.0). The internet ushered in online banking (3.0). Smartphones enabled mobile-first, platform ecosystems (4.0). But each stage, while technologically impressive, remained shaped by a legacy mindset: products first, customers second.
As I explore in Brilliant Data, Banking 5.0 is different. It represents the fusion of AI’s predictive power with human-centered design. It asks not only “What can we automate?” but “What outcomes do people need?” It demands new financial business models that merge AI with behavioral economics. AI identifies patterns and predicts outcomes with precision; behavioral economics explains why people decide the way they do, shaped by biases, heuristics and emotion.
Where one discipline provides foresight, the other provides context. Together, they allow finance to move from reactive to predictive, from generic to personalized, and from transactional to preventative – at scale.
When combined, these disciplines give us a powerful toolkit for designing financial services that anticipate needs, align with human behavior, and foster trust.
This duality – machines plus meaning – is what sets Banking 5.0 apart. It is not just about efficiency. It is about empathy at scale, delivered through new business models that scale like super apps, digital assistants or digital guardian angels.
Applications for brilliant data
Financial institutions sit on oceans of behavioral data: payment rhythms, app taps and swipes, pauses and hesitations. Yet for years, much of this was treated as a compliance artifact rather than a source of value.
In our fintech work (and my PhD research), we found that behavioral intelligence consistently outperformed legacy proxies. When we complemented traditional variables with psychometrics and behavioral signals, real differences emerged (see box, p62). They indicate the potential value that can be realized by utilizing the brilliant data that has been so far neglected.
Banking 5.0 is not a theoretical future. It is already reshaping financial services across multiple dimensions.
Credit scoring: expanding opportunity
Traditional scorecards exclude the very people building the future: students, gig workers, immigrants, sole traders, and many women-led and first-time businesses. Thin-file segments total about 2 billion people. If behavioral models help us responsibly approve 30% more of them, that’s about 600 million new participants, and an estimated $2.5 trillion reduction in the access gap – without raising risk tolerance.
Risk management: from early signals to early action
AI can scan millions of transactions in real time. Add behavioral cues – timing shifts, stress indicators – and banks can spot vulnerabilities before arrears, enabling proactive restructuring, hardship support or fraud interventions. This is prudence with empathy. This enables proactive outreach, protecting both customers and institutions.
Customer engagement: nudges that build resilient habits
Instead of generic product pushes, use behavioral design to deliver goal-aligned nudges: save when income spikes, calm over-spend surges, pre-commit to bill smoothing, micro-invest windfalls. Adoption rises when advice feels personal, timely and non-judgmental.
Inclusion as a growth strategy
Next-gen customers move to apps. Migrant workers trust lower-cost remittances when transparency is high and flows are explainable. Low-income customers build buffers with round-up savings. SMEs access dynamic, cash flow-based credit that flexes with seasonality. Inclusion ceases to be charity; it becomes profitable growth expansion.
Retention: loyalty through empathy
Customers remain loyal to institutions that understand them. AI-driven personalization – financial health scores, dynamic limit-setting, proactive alerts, relevant offers – shows empathy. Trust, once lost, is hard to rebuild. Banking 5.0 helps build lifetime value from the start.
These are not hypothetical possibilities. They are the lessons of lived innovation – successes, failures, and the surprising discoveries that shaped both my companies.
Challenges and ethical considerations
Powerful tools demand responsible use, and Banking 5.0 raises ethical questions that leaders cannot ignore. Behavioral insights can easily cross the line from helpful nudges to manipulation – pushing higher-margin products that serve the institution more than the customer. The boundary between guidance and exploitation is thin, which is why safeguards matter: radical transparency about intent and impact, independent oversight, and a genuine customer-first ethos.
Trust also hinges on data stewardship. Financial data is among the most sensitive information people possess, so institutions must act as custodians, not extractors. Compliance with data privacy regulations like Europe’s GDPR or California’s Consumer Privacy and Privacy Rights Acts (CCPA and CPRA) is only the baseline. True differentiation comes from clear, comprehensible practices, and demonstrable customer benefit.
Finally, human oversight remains essential. AI models are powerful but imperfect; they can misread behavior or reproduce bias. Decisions that affect lives – loan approvals, fraud alerts, investment advice – must include human judgment. AI should augment, never replace, human responsibility. Handled with care, these challenges become opportunities for leadership; mishandled, they erode the very trust on which banking depends.
Strategy first: the asymmetric advantages of Banking 5.0
The next question is strategic: why does Banking 5.0 win, commercially and ethically, without asking leaders to take more risk? The answer lies in four asymmetric advantages that compound over time. They come from seeing behavior more clearly – not from loosening standards.
- New addressable market, same risk appetite Behavioral models surface good risk that legacy ratios miss – thin-file customers, first-time entrepreneurs, seasonal SMEs. You expand profitable inclusion without relaxing thresholds, adding growth by better discrimination rather than greater exposure.
- Bending the opex curve Human-centered AI democratizes banking guidance. As advice shifts from human-only to human plus AI, unit economics improve: always-on coaching, consistent quality, and lower cost to serve – especially in segments incumbents struggle to reach. That has a direct impact on operating expenditures (opex).
- Compounding trust Visible fairness, clear explanations and proactive support translate into loyalty, referrals and higher lifetime value. In a commodity market, trust becomes the moat that spreadsheets can’t copy – and it gets stronger with every transparent decision.
- Speed with integrity Behavioral models can run as additive or threshold scorecards alongside existing systems, so you can move fast without betting the bank. You learn in parallel, keep humans in the loop, and scale what works – safely.
Together, these four advantages turn inclusion into a growth engine and prudence into a differentiator – setting up the leadership moves that follow.
Leadership imperatives for transformation
Banking 5.0 is as much about leadership as it is about technology. My own path from corporate executive to entrepreneur taught me that strategic courage, an innovative culture, and ethical responsibility are the real levers of change. This means shifting our gaze from products to outcomes: boards and executives must measure success not by the number of loans booked but by the number of lives improved.
It also means moving from silos to squads. These capabilities cannot live at the periphery: data scientists, behavioral economists, designers, product managers, risk and compliance need to work side by side, so that academic depth meets commercial pragmatism.
Leadership in this era also requires a longer horizon. For me, it is not about the next quarter but the next generation. As a father of four, I’m acutely aware that we are not building financial systems for today alone. We are building for a world shaped by climate change, inequality, and rapid digital transformation, and that demands designs that serve not only shareholders but society.
These imperatives are not abstract. They are the lessons I wished I could have applied more fully inside large institutions – and the ones I have tested, sometimes painfully, as an entrepreneur. Taken together, they describe the essence of leadership in the Banking 5.0 era.
Toward a more inclusive financial future
Banking 5.0 is not only about faster transactions or lower costs. It is about reimagining finance as a driver of value, trust, and inclusion by blending AI’s predictive intelligence with behavioral economics’ human insight. Done well, it delivers truly personalized services tailored to each person’s context, provides preventative support that addresses risks before they escalate, and builds inclusive systems that bring underserved communities into the financial mainstream. This is finance not as a gatekeeper, but as an enabler; not as a product machine, but as a human partner.
For customers, it means advice that’s accessible, timely, and humane; private-banking-grade coaching without the private-banking price tag. For SMEs, character and cash flow matter again; viable businesses aren’t turned away by a ratio logic rooted in the 1920s. For embedded finance players, it means opening a profitable new frontier, while cutting avoidable losses. And for society, it means that financial capability increases, defaults fall, entrepreneurship rises, and trust returns. Critically, banking 5.0 isn’t about loosening standards. It’s about seeing people more clearly.
Finance can indeed be more human, more inclusive and more just. But it requires leaders with the courage to innovate beyond products, to think beyond quarters, and to design for the next generation. Finance should be the engine that powers possibility. The numbers say the market is ready; the technology says the tools are ready. The only open question is leadership. Start now.
