What banks keep getting wrong about AI

Welcome to Tech in Asia’s daily newsletter, your essential dose of Asia’s tech and startup buzz. Not yet a subscriber? Register here. Got a story tip? Send it to editors@techinasia.com.

In focus

• ​​The challenge of integrating AI into private banking persists
• ​​AI is reshaping BNPL firms’ operations across Southeast Asia
• PH fintech startup Salmon raises US$100 million

Hello reader,

There’s a tell when someone hasn’t fully bought into using AI at work. They’ll say AI’s output isn’t worth the time it takes to check it.

You hear this a lot from writers, developers, teachers, and analysts. The complaint is always some version of the same thing: By the time they’ve corrected what AI got wrong, they realize they could’ve just done it themselves.

It’s a fair frustration built on a lack of trust and a flawed approach of using AI just to optimize efficiency. And efficiency, it turns out, is a weak reason to adopt any new tool.

Banks are learning this the hard way. Most of them have poured money into AI, only to see little in return. The reasons, as our top story today shows, don’t have much to do with the tech itself.

Most banks deploy AI in their private banking divisions by putting it in the background and letting it automate prep work, generate pre-meeting briefs, and cut admin time. But revenue in private banking is made during client conversations – not before them – so getting AI to help influence client decisions in real time is the better option.

Getting there, however, requires banks fixing their data foundation so it’s less fragmented across divisions. They will also need to rebuild their workflows and rethink how they handle compliance. Unfortunately, none of these came with the tools they bought.

This is maybe the best way to think about AI adoption, in banking or anywhere else: The tech is rarely the hard part.

Glenn Kaonang, journalist

Top Story

Banks pour billions into AI, but most see no ROI. Here’s why

Image credit: Timmy Loen

A report drawing on executive interviews reveals just how steep the learning curve for AI in banking is. While a bank can get a new AI model running in just three months, it often takes nine months before relationship managers trust it enough to use it.

A visual story you shouldn’t miss

Trending news

Registration is now open for WorldQuant Brain’s International Quant Championship