Nubank elevates customer experiences with OpenAI

Nubank has never been shy about calling itself a tech company that happens to do banking, and its latest move leans hard into that identity. The Brazilian digital bank, which now serves well over 100 million customers across Brazil, Mexico, and Colombia, has been working with OpenAI to rebuild how its support systems talk to people. Instead of routing every question through rigid scripts or long queues, the company is using large language models to understand context, pull account details, and respond in a way that feels less like talking to a call center and more like a competent human who already knows your history.

What makes this interesting isn't the novelty of a bank using AI chat — plenty of banks have bolted a chatbot onto their app and called it innovation. It's the scale. Nubank operates in a region where traditional banking infrastructure has historically been slow, expensive, and often hostile to lower-income customers. The company built its entire business on stripping out friction, and customer support has always been part of that pitch. Layering OpenAI's models on top of that foundation means the bank can, in theory, resolve disputes, explain fees, or walk someone through a loan application without a human agent ever getting involved, freeing up staff for the genuinely thorny cases.

There's also a language and cultural angle that's easy to miss if you're not paying attention to fintech outside the US and Europe. Serving Portuguese-speaking Brazilians and Spanish-speaking Mexicans and Colombians with a single AI-driven support layer, tuned to local dialects and financial norms, is a harder problem than it sounds. Nubank's bet is that a capable model, given the right guardrails and data access, can flex across those markets more cheaply than hiring and training thousands of region-specific agents.

None of this is charity, of course. Faster resolution times and fewer human agents translate directly into lower operating costs, which matters a lot for a bank that trades on razor-thin margins per customer and makes its money on volume. OpenAI, meanwhile, gets another marquee financial-services customer to point to when it argues that its models are ready for regulated, high-stakes environments — not just drafting emails or summarizing documents.