Paulo Victor Leite Lima Gomes

Posted on Jul 14

why Mal may beat Revolut: the first AI-native Islamic bank is not just adding a chatbot

#ai #fintech #banking #product

I like Revolut.

That is probably the right place to start, because this is not a "Revolut is bad" article.

Revolut is one of the best consumer fintech products ever built. It has distribution, brand, UX, payments, cards, trading, crypto, savings, business accounts, travel features, and the kind of product velocity most banks still pretend is impossible.

So when people ask whether a new digital bank can beat Revolut, the lazy answer is no.

Scale wins.

Licenses win.

Brand wins.

But the more interesting question is not whether a new bank can copy Revolut faster than Revolut can expand. That game is mostly over. The interesting question is whether a new bank can play a different game.

That is why Mal is interesting.

Mal describes itself as "the first AI-native Islamic digital financial platform." Fine, that sounds like positioning. Every fintech deck now has some version of "AI-native" in it.

The difference is the wedge.

Mal is not just saying: we have a banking app and now we added a chatbot.

The more ambitious version is: we are designing the financial workflow itself around intelligence, values, compliance, and customer outcomes from day one.

That is a much more dangerous idea.

the AI banking race is not about chat

The most boring version of AI banking is a support bot.

You ask why your transfer failed. The bot says it is sorry. You ask again. The bot explains the same thing with slightly better grammar. Eventually you type "human" twelve times and remember why branches existed.

That is not the race.

The race is not even who generates more code internally. That matters for cost and speed, but customers do not care whether your backend was written by a staff engineer, Copilot, Codex, or a very tired team in a room with bad lighting.

Customers care about outcomes.

Can I understand my money?

Can I make a better decision?

Can I trust the product?

Can I get through onboarding without feeling like I am applying for a mortgage in 1998?

Can the bank explain what it is doing?

Can it prevent me from making a stupid mistake without becoming paternalistic?

Can it reduce fraud and friction at the same time?

That is where AI becomes interesting in banking. Not as a chat window. As a workflow layer.

Mal's wedge is not just AI

Mal's wedge is AI plus Islamic finance.

That combination matters.

Islamic finance is a massive market. Finews described it as a USD 7 trillion opportunity and reported that Mal raised more than USD 230 million in what it described as the largest seed round in the Middle East and Africa. Fintech News UAE later reported that Mal received in-principle approval from the Central Bank of the UAE to establish an AI-native Islamic digital bank.

The wording matters: in-principle approval is not the same as a fully launched bank. Mal still has to prove licensing, governance, product quality, uptime, customer trust, and the boring operational discipline that banking requires.

But the market problem is real.

Islamic finance is not just "normal finance with a religious checkbox." At its best, it is a values-based financial system with real constraints around interest, excessive uncertainty, harmful activities, and the relationship between money and society.

That creates a different trust surface.

Revolut competes on speed, breadth, cost, FX, convenience, and product density.

Mal can compete on something harder to copy: a bank that makes the values of the product visible in the product.

Not a PDF.

Not a compliance page nobody reads.

Not a generic "we care about ethics" paragraph.

The product itself.

what AI-native could actually mean

The phrase "AI-native" is dangerously close to becoming meaningless, so let me define what I would want it to mean in a bank.

AI-native does not mean every screen has a sparkle icon.

It means the bank is designed so intelligence is part of the operating model.

Onboarding should be adaptive. If the customer already gave enough information, stop asking for more. If the risk is higher, ask better questions, not just more questions. If the customer does not understand a requirement, explain it in context.

Customer support should know the state of the account, the product, the attempted action, the applicable policy, and the likely next step before escalation. Not because it is creepy, but because the bank already has the context. Pretending not to have it is bad service.

Budgeting should not just categorize spend. It should understand goals, constraints, recurring obligations, upcoming travel, family context, and risk. It should say: this is the trade-off, not just this is the chart.

Shariah and ethical assurance should be explainable. If a product is compliant, explain why. If a transaction, merchant, or investment is questionable, explain the reasoning. If there is uncertainty, show it. In finance, "trust us" is not a product feature.

Risk and fraud should become less binary. A good AI-native workflow should reduce false positives, ask for the right confirmation at the right moment, and protect customers without turning every unusual action into a blocked account.

That is the real game.

Not "AI writes our code."

"AI changes how the bank works."

why the interview made this clearer

In Abdallah Abu-Sheikh's interview, the part that stood out to me was not a model demo or a technical claim.

It was the way he talked about problem statements, community, ethics, and building from the region instead of waiting for somebody else to solve the problem.

That is important because a bank is not a neutral container for features. A bank has an opinion about risk. It has an opinion about who gets access. It has an opinion about what a good financial decision looks like. Most banks just hide those opinions behind policy.

If Mal is serious, the opportunity is to make those opinions explicit and useful.

An Islamic digital bank should not just say "this product is Shariah-compliant."

It should help the customer understand the financial, ethical, and practical consequences of the decision in front of them.

That is much closer to a workflow problem than a chatbot problem.

where Revolut is strong

To be fair, Revolut is not standing still.

Revolut has enormous advantages. It has tens of millions of customers, a global brand, product breadth, financial infrastructure, regulatory experience, and a habit of shipping fast.

Its AI work is also real. Revolut's own AI Assistants notice describes AI Support and an AI Revolut Assistant that can help with things like support, spending analysis, budgeting and travel information, and factual product or investment information. The same notice is careful about scope: it says the assistant does not provide financial advice or execute financial orders.

That is a sensible boundary.

It is also the point.

Publicly, Revolut's AI story still reads mostly like an assistant added to a large existing product surface.

That can be useful. It can reduce support load. It can make the app easier to navigate. It can summarize information.

But it is not the same as designing the bank itself around AI-assisted workflows, explainable policy, values-aware guidance, and adaptive operations from day one.

Revolut has the advantage of scale.

Mal has the advantage of a blank sheet.

The blank sheet only matters if they use it properly.

the hard part

This is where the hype has to stop.

Mal does not win because it says "AI-native."

Mal does not win because Islamic finance is large.

Mal does not win because the seed round was huge.

It has to execute.

It has to be licensed properly. It has to earn trust. It has to make the app reliable. It has to make the ethical layer visible without making the product feel slow or preachy. It has to avoid the classic fintech trap of beautiful onboarding followed by fragile operations.

And it has to be honest about AI.

AI in banking cannot be a vibes engine. It needs governance, auditability, fallback paths, human escalation, model risk management, privacy boundaries, and clear customer responsibility.

If the AI gives an explanation, the bank needs to know where that explanation came from.

If the AI flags a transaction, the bank needs to know why.

If the AI helps a customer choose, the bank needs to make clear what is guidance, what is information, what is advice, and what is the customer's decision.

That is not easy.

But that is exactly why it could be a moat.

my bet

If Mal beats Revolut in this race, it will not be because it is "more Islamic."

It will not be because it has "more AI."

It will be because it turns a clear community, a real values framework, and a modern AI workflow into a better banking operating model.

Revolut made banking feel like software.

Mal's opportunity is to make banking feel like an intelligent, ethical workflow that actually helps the customer make better decisions.

That is a very different promise.

And if they execute, it is a promise Revolut cannot simply copy by adding another assistant tab.

sources

To test my projects, I use Railway. If you want $20 USD to get started, use this link.

DEV Community