BFSI Is About to Scale AI. Most of Them Aren't Ready.

By April of this year, Uber's CTO Praveen Neppalli Naga had a problem. His engineers had embraced AI coding tools with exactly the kind of enthusiasm you would expect from a company that prides itself on moving fast. Claude Code, Cursor, the works. Productivity was up. The tools were good. Really good.

And then someone looked at the bill.

Uber had burned through its entire 2026 AI budget by April. Four months into the year, with eight months still to go. Individual engineers were running up between $500 and $2,000 a month in API costs. Nobody had modeled what that looked like across thousands of engineers. There were no real-time caps. No governance layer watching what was happening. Just a lot of very productive engineers and a budget that quietly disappeared while everyone was celebrating the productivity gains.

"I'm back to the drawing board," Naga told The Information, "because the budget I thought I would need is blown away already."

Uber will be fine. They have the resources to absorb a budget shock and redesign their approach. Most banks and financial services firms I talk to are not Uber. And the version of this story that plays out in a regulated financial institution is not just expensive. It is a compliance event.

This Is Not a Technology Problem

My background is in business development and digital transformation. Together with partners like TribalScale, I have sat across the table from CIOs and CTOs at some of the largest financial services firms in North America. I have seen what they invest in, what they are trying to solve, and more instructively, what happens six months after the enthusiasm of a new program fades and the hard work of actually operationalizing it begins.