The Cultural API

There’s a quiet arrogance embedded in global finance — a belief that compliance is universal, that identity is binary, that “knowing your customer” means checking a few boxes that were drawn up in London or New York. The result? A sprawling, global system that claims to know its customers but in truth doesn’t understand them at all.

I’ve been around the world enough to see this breakdown in real terms. A startup founder in Nairobi. A medical researcher in Lahore. A business family in Muscat. All entirely legitimate, all frozen out of financial systems because their way of demonstrating identity doesn’t match the narrow template baked into traditional KYC frameworks.

The Western model of identity is rooted in documents. Proof of address. Tax registration. Utility bills. But what if those things don’t exist in the way you expect? What if trust is built through tribal lineage, local networks, or decades-long business relationships that are entirely offline?

What happens is exclusion. Not because the person is untrustworthy, but because the system has no way to interpret trust that doesn’t speak its language.

This is why I believe strongly compliance is no longer just a regulatory function — it’s a cultural problem. And like all cultural problems, it requires new lenses.

We need to build a new kind of compliance logic — one that doesn’t just verify, but understands. Think of it as a cultural API — a layer that sits between local context and global systems, translating behaviours, norms, and signals into language that the financial system can comprehend. Not by flattening the culture out of them, but by making the system more literate.

This isn’t science fiction. We have the tools now. AI is already being used to analyse complex behavioural data. Machine learning models can already detect subtle patterns. But we’re using that power to chase fraud and flag anomalies — not to build a richer picture of trust.

Imagine an AI system that can recognise the nuance in how a person interacts with digital platforms, cross-reference that with regional behavioural norms, and map it against known-good actors in a local trust web. Not to replace documentation, but to complement it. To see the full human, not just the paperwork.

This would revolutionise what KYC actually means. Instead of reducing someone to a document trail, we’d elevate them to a trust signature. Dynamic. Multi-dimensional. Culturally aware.

And this shift would do more than just include more people — it would reduce risk. Because let’s be honest: the current system isn’t just exclusive. It’s ineffective. Criminals know how to game the forms. Good people often don’t even get a chance to fill them in.

It’s not just the Global South that suffers. The entire system becomes brittle when its only inputs are checkbox data and PDF scans. If you want resilience, you need richer signals. And those signals come from culture.

I’ve long believed that inclusion won’t come from loosening standards. It will come from upgrading intelligence. From moving away from static compliance and toward interpretive understanding. From systems that don’t just say “yes” or “no,” but ask, “why?” and “in what context?”

To build that future, we need engineers who speak anthropology. Compliance officers who think like ethnographers. Regulators who understand that fairness is not the same as uniformity. AI gives us the ability to pull together these currently dissonant disciplines.

Because knowing your customer — and I mean really knowing them not “KYC’ing them” — isn’t about standardisation. It’s about translation.

And if we don’t learn how to speak the world’s many languages of trust, we’ll continue to mistake good people for bad risks — and miss the greatest opportunity for systemic reinvention that this century has to offer.