AI is helping revolutionise industries everywhere, but it can also have benefits for customers and consumers.

In the banking sector, AI is set to bring significant changes, with smart tools able to help customers overcome the biggest limitation of AI prompting: knowing what to ask.

The last 18 months have seen Starling launch a suite of customer-facing generative and agentic AI tools, and we spoke to its deputy CIO Frederic Laurent to find out more.

• Most banks are using AI behind the scenes, but Starling is really dialling up its customer facing tools. What are you hoping to achieve here?

When we received our banking licence ten years ago, we were part of a breed of challenger banks that pioneered app based banking. We’ve applied our tech expertise to solve so many customer problems in that time, from tools that help customers control their gambling; lock their card if they lose it; or get spending insights about where their money really goes. Many are now at play at traditional banks today.

But there are so many more problems that technology hasn’t been able to address yet; the UK’s financial literacy levels are low, many people don’t feel fully in control of their money, and scams are on the rise. Those are just a few of the challenges we think AI can tackle.

Granted, some of these challenges like fighting fraud can be addressed with back office AI applications. But it would be a shame to only use this technology behind the scenes, when it could have far greater impact in customers’ hands. We therefore knew that when AI hit the public arena, and was no longer the reserve of technology teams, we should be ready to roll out customer facing AI tools of our own.

We saw the launch of customer facing LLMs as that trigger point and, using Google’s Gemini models, we’ve rolled out a number of tools that were first to market. Our Spending Insights tool uses Gen AI to help customers interrogate their spending using natural language; finding out how much you spent on takeaways last month can make you seriously reconsider your budget. Meanwhile, our Scam Intelligence tool educates our customers on the warning signs of scams; it increased the rate that customers cancel marketplace payments by 300%, and was welcomed by the UK Fraud Minister.

• You moved beyond Gen AI to agentic, launching an agentic AI financial assistant earlier this year. What does it do and what impact is it having?

If Gen AI does the thinking, then agentic AI does the doing, and Starling Assistant leverages both capabilities to respond to customer queries and carry out actions on their behalf.

It helps customers feel more in control of their money, whether it’s creating budgets for them, setting up saving plans, opening Spaces or running through a list of their direct debits - all by responding to natural language prompts.

Another game changer is the new interface it brings to banking. Customers are no longer confined to buttons and app screens - they’re presented with a blank page and an opportunity to share the problems they want the assistant to solve.

We launched this expecting our younger or more tech savvy customers to use it most frequently, but actually it’s older customers or those who are new to the bank that have embraced it most. For many, asking simple questions is the easiest way to manage their money.

• Is adoption as strong as it could be? What can you do to encourage greater sign-ups?

Not yet - usage mirrors what we’re seeing with general public AI models in that some people love AI, some dabble and some don’t see the full potential.

We’ve also realised that some customers suffer from writers’ block. When faced with an AI assistant that has seemingly limitless opportunities, they can be paralysed by indecision or lack of imagination on what commands to give.

We’ve launched our new smart tools programme to give our customers more inspiration. Think of it like a set of predefined prompts that show customers exactly how AI can help them, from preparing SMEs for Making Tax Digital, helping university students maximise their maintenance loan, or helping customers set up the right scam controls for their needs.

We’ll be rolling out new smart tools weekly to keep the inspiration going, and we’re encouraging customers to share their own requests with us too. Soon, we’ll be allowing customers to build their own smart tools, save them to the assistant and create an AI banking experience that’s entirely bespoke to their needs.

• What do you think comes next in this field?

We’re looking forward to seeing even more customers adopt our tools, and more banks offer similar capabilities. There’s no question in my mind that AI is the future of money management and, collectively, the banks can drive that behaviour change.

At Starling, we’ll keep the smart tool engine running and will continue to address customer problems as we have always done. Another frontier for us is agentic payments and commerce; watch this space.

Mike Moore is Deputy Editor at TechRadar Pro. He has worked as a B2B and B2C tech journalist for over a decade, including at one of the UK's leading national newspapers and fellow Future title ITProPortal. When he's not keeping track of all the latest enterprise and workplace trends, he can most likely be found watching, following or taking part in some kind of sport.