Julián Colombo, CEO of N5, was interviewed by Pool Económico GB to discuss one of the deepest shifts underway in the financial industry: the move from manual transfers to automated, event-driven transactions powered by predictive models.
In the interview, Colombo explains how artificial intelligence has stopped being an innovation experiment and become the operating engine of banks and fintechs, and why the sector’s next competitive edge won’t belong to whoever has the most technology, but to whoever turns it into trust, efficiency, and a better customer experience.
Here are some of the most relevant excerpts from the conversation.
On the safety of “invisible money”
“It’s not about the user losing control, but about technology executing decisions the person has already defined: rules, limits, permissions, and conditions. AI shouldn’t ‘do whatever it wants’ — it should operate within a secure, traceable, auditable framework.”
On conditional payments
“A person could set up their rent to be paid only once their salary has been received and their balance stays above a certain threshold. For an SMB, a payment to a supplier could be triggered only once the goods have been delivered, the invoice validated, and cash flow allows it. The payment stops being an isolated action and becomes part of an intelligent process.”
On banks’ legacy systems
“Banks have a huge advantage: trust, regulation, customers, data, and scale. […] The risk isn’t disappearing — it’s losing relevance in the customer experience. […] The old systems don’t need to be replaced wholesale; you build around them: connecting via APIs, rules engines, and AI-, BPM-, or RPA-powered business layers.”
On accuracy and hallucinations in financial AI
“We work to keep the AI suite our clients use at a 0.8% hallucination rate, when chatbots on average range between 3% and 27%.”
On the next step for banking leaders
“Stop seeing AI as an innovation experiment and put it at the center of business strategy. The first step is mapping customer friction points and operating costs, identifying which data and systems need to be connected, and choosing use cases with real impact and the ability to scale. The advantage won’t belong to whoever has the most technology, but to whoever turns it into trust, efficiency, and a better experience.”
Want to read the full interview? Julián Colombo also discusses who’s responsible when AI makes a mistake, how to make an algorithm explainable to a regulator, and where the global $97 billion investment in AI for financial services is headed.
