Over the previous month, Nubank has been making headlines, and it appears to be like more and more doubtless that the Brazilian digital banking large has ambitions past Latin America.
In September 2026, Nubank moved nearer to coming into the US retail banking market via a partnership with Lead Bank. With round 140 million clients globally, it’s already the world’s largest digital banking platform by buyer base, with roughly twice as many purchasers as Revolut.
That’s a powerful place to begin however cracking the US is one other matter.
Nubank has additionally acquired conditional approval from the OCC for a US nationwide financial institution constitution, with the purpose of securing full approval by 2027. A full constitution might give the firm larger management over its US operations and a stronger basis for competing in one in all the world’s most profitable banking markets.
Then got here the Monzo rumours.
Reports steered Nubank might purchase Monzo as a part of its European enlargement technique. For Nubank, that might imply a longtime foothold in the UK. For Monzo, it might carry the monetary backing and worldwide attain to revisit its personal US ambitions, the place earlier enlargement efforts struggled to achieve traction.
There’s an attention-grabbing irony right here. Monzo and Revolut have been each based in 2015, with Monzo getting began round 5 months earlier. Monzo additionally secured its UK banking licence years earlier than Revolut. Yet Revolut has since constructed a wider worldwide enterprise, whereas Monzo stays extra carefully tied to its house market.
Could Nubank be the worldwide heavyweight Monzo wants? Perhaps. But Nubank was fast to dismiss the acquisition hypothesis, saying it had no plans to purchase the British digital financial institution.
I can perceive why.
Nubank has constructed its buyer base by specializing in markets it is aware of. Buying its manner into Europe via Monzo would possibly speed up enlargement, however it might additionally imply taking up one other firm’s technique, regulatory challenges and development ambitions. An acquisition brings scale, but it surely doesn’t robotically carry success, notably when Nubank is already working to ascertain itself in the US.
So the place does that depart Nubank?
It has the buyer base, the model recognition and the ambition. But worldwide enlargement and replicating success are two very various things.
The US is a crowded market, with established banks, digital challengers and fee firms competing for purchasers. Acquiring them will be costly. Getting folks to enroll is one factor; convincing them to maneuver their cash, make Nubank their main banking supplier and keep for the long run is way tougher.
Nubank has already proved it will probably scale. What it hasn’t but proved is that its mannequin can journey efficiently right into a market as aggressive and complicated as the US.
That’s the actual take a look at. Cracking America isn’t merely about including thousands and thousands of consumers. It’s about turning that development right into a sustainable, worthwhile enterprise. And 140 million clients elsewhere received’t assure that occurs.
Jamie Dimon: Get Out of the Way? The Battle for Agentic Payments
I’m totally having fun with all the chatter round AI brokers and funds. The business is imagining a world the place brokers can store, evaluate costs, handle subscriptions and make funds on our behalf. It sounds handy, and the business alternative is gigantic.
But there’s a query that deserves simply as a lot consideration as the expertise itself: when an AI agent will get one thing improper, who’s liable?
The concern isn’t simply that an agent would possibly go off script or be manipulated into making an unauthorised transaction. It’s that accountability might turn into more and more troublesome to pin down when the buyer, financial institution, fee supplier and firm constructing the agent are all concerned.
The US has spent years constructing safeguards round buyer knowledge, privateness, fraud prevention and entry to monetary accounts. Banks make investments closely in id verification, transaction monitoring and compliance. Now we’re speaking about giving AI brokers the capability to behave on a buyer’s behalf.
The potential is difficult to disregard. An agent might evaluate monetary merchandise, negotiate payments, handle subscriptions and discover the greatest method to pay for one thing. Customers would possibly not often must open a banking app.
But comfort comes with a catch. If an agent can entry your account and provoke funds, whose pursuits is it serving? The buyer, the financial institution, the expertise firm or the service provider making an attempt to make a sale? And if the agent makes a mistake, who picks up the invoice?
Banks can’t merely blame the algorithm. Equally, expertise firms can’t anticipate banks to soak up each loss whereas they management the buyer interface.
Banks Are Building Their Own. Is That the Right Move?
Banks are understandably eager to develop their very own AI brokers. They have already got the buyer relationships, monetary knowledge, regulatory infrastructure and, crucially, entry to the cash.
Building their very own brokers provides them larger management over buyer data and what these brokers can do. It additionally reduces the danger of expertise firms inserting themselves between banks and their clients.
But is constructing every little thing in-house actually the proper technique?
Banks perceive danger, regulation and monetary providers. That doesn’t robotically make them the greatest at constructing client AI merchandise or creating the agent ecosystems clients will need to use.
There’s additionally the danger of fragmentation. If each financial institution develops its personal closed system, we might find yourself with dozens of competing brokers that battle to work throughout banks, retailers and fee networks. That hardly feels like the seamless expertise the business is promising.
And let’s be sincere, banks have spent years making an attempt to make their apps simpler to make use of. Now they’re being requested to construct programs that perceive what clients need, make choices and probably transfer cash with restricted human intervention.
That’s fairly a leap.
The various isn’t essentially higher. If impartial expertise firms turn into the essential interface between clients and their banks, monetary establishments danger being pushed into the background. They would nonetheless present the accounts and course of the transactions, however another person might management the buyer relationship.
For banks, that’s a critical concern.
The Real Battle Is Over Control
Imagine asking an AI assistant to seek out the greatest bank card, evaluate financial savings accounts or handle your payments. Rather than opening a number of banking apps, you inform the agent what you need and let it do the work.
The agent turns into the place to begin for the determination. It might affect which monetary product you select, which financial institution you employ and the way you pay. The financial institution would possibly nonetheless present the underlying service, but it surely might have far much less affect over the buyer’s alternative.
That’s a big shift in the stability of energy.
Payment networks, retailers and expertise firms even have a stake on this new layer of commerce. Whoever controls the agent might affect the place transactions occur, how funds are routed and which suppliers are chosen.
For years, banks and fintechs have competed to personal the buyer expertise. Agentic AI might put that relationship in the fingers of a very totally different set of gamers.
This is not only a dialog about chatbots. It’s about who will get to take a seat between the buyer and their cash, and who captures the worth that comes with it.
The Liability Question
Before agentic funds turn into mainstream, the business wants clearer solutions on accountability and management.
Who authorises a fee? Can an agent be restricted to a specific quantity, service provider or goal? What occurs whether it is compromised? And who’s accountable when a transaction goes improper?
These questions turn into extra sophisticated when an agent operates throughout a number of suppliers, every with its personal programs and obligations.
Existing fee protections nonetheless matter, however their utility to agent-driven transactions wants cautious consideration. Giving an agent permission to make funds doesn’t essentially imply the buyer supposed to authorise each transaction it would provoke.
The business wants to ascertain clear guidelines round consent, permissions and accountability earlier than clients will be anticipated to belief brokers with more and more consequential monetary choices.
Because if customers don’t know who’s accountable when one thing goes improper, comfort will solely take the business thus far.
The Next Battle in Payments
I can see why banks need to construct their very own brokers. They have an excessive amount of at stake to let another person take management of the buyer relationship with out a combat.
But constructing every little thing in-house might not be the reply both. Agentic commerce might want to work throughout banks, retailers, fee networks and expertise suppliers. Some stage of cooperation might be unavoidable.
The problem is discovering the stability between openness and management: giving brokers sufficient entry to be helpful with out compromising buyer safety or leaving accountability unclear.
Financial providers have spent years making an attempt to make funds sooner, less complicated and safer. Agentic AI might take comfort to a different stage, but it surely additionally raises new questions on consent, accountability and who finally controls the transaction.
Banks shouldn’t ignore the alternative. But neither ought to they assume that constructing their very own brokers will robotically defend their place.
Because the subsequent battle in funds might not be about who strikes the cash.
It could also be about who decides the place the cash goes.
AI stage 2 of 5: drafted by our AI editorial assistant from supply materials our editor selected; fact-checked, edited and signed off by Mark Walker, Editorial Director. What the ranges imply
