A tale of two banks, simplification, and AI - NatWest Group’s story

NatWest Group is one of the top five banks in the UK, whose brands including National Westminster Bank, Royal Bank of Scotland, Ulster Bank, and Coutts, service over 19 million customers. It’s a complex operation with around 60,000 employees in total and it comes as no surprise that a top priority for Group CEO Paul Thwaite is tech-enabled simplification:

We have already made significant progress simplifying our systems and reducing duplication. For example, we de-commissioned 200 business applications across the group last year, and we successfully migrated one million customers from Sainsbury's Bank, covering multiple products.

It will come as another non-surprise to find out that AI features heavily in future planning for the banking group:

It's already affecting the sector. We've embraced it - from a colleague perspective and a customer perspective. I think it's going to change how customers engage with us or how they find us and discover us. I think the winners here will be those who've got significant sized customer bases, 20 million for us, long-standing relationships data, so you can bring products, propositions, whether directly to your own channels or through other channels. I think that is going to be successful. And we're very thoughtful about that in terms of how we're building our capabilities. 

Successes

Thwaite is able to point to early AI successes:

Deployment of AI is not only helping us to automate routine work, such as call summarization, it is also helping our coders to be more productive. Over 12,000 software engineers are now able to use AI assistance to generate code. This transformation has enabled us to improve the deployment frequency of updates across the group by more than 4x since 2021 and more than trebled the new features on our commercial banking digital platform Bankline.

The productivity of software engineering is, according to Thwaite, “definitely a topic du jour” for the NatWest Group management team:

It's pretty obvious the AI developments have been transformational for us. All our engineering and coding teams have got access to AI tools. As you alluded to, we have around 12,000 engineers and that's been increasing over a number of years...We've got a couple of quite exciting pilots running in two  of our businesses in our international business and also in our financial crime area, where we’re doing fully agentic ‘press play’ software, and that's actually delivering 10x productivity gains. That's where you've got agentic workflows, autonomous agents, their planning, building code, testing code, but obviously then overseen in a responsible way by human.

But now the Group is at a situation where circa 35% of the code is written by AI. What does that mean for the human engineers prospects within NatWest? Thwaite is non-committal:

I think, over time, there will be choices around how you use that capacity. It's still an evolving picture...this space is, I think, exploding pretty quickly. And I think it's inevitable there'll be a change both in the profile of, let's call it, engineers in terms of the activities that they do. And then I think there'll be some choices about how you capture that productivity benefit to capture some of it to go faster, deliver more products and services to your clients and enhance the customer proposition.

He adds:

We also see opportunities for productivity and efficiency. I think all of the things being equal, that's a reasonable expectation over the short to medium term, that there'll be some productivity and efficiency opportunities moving forward... [We are] very excited by the work that's going on there but we are very mindful that we're a regulated industry, and we're doing it in a very responsible and thoughtful way. 

M&A and partnering

The firm has also been partnering to bring in AI expertise and resources, most notably via its collaboration with Amazon Web Services around data, analytic and AI capabilities. This is intended to provide a single view of each customer's relationship with the bank as well as the tools to analyze data and enrich customer understanding.

Last week NatWest upped its Wealth Management footprint by announcing a £2.7 billion Evelyn Partner, a leading UK wealth manager with more than 180 years of heritage overseeing £69 billion of Assets Under Management and Administration. This is a big deal for the Group, says Thwaite:

In terms of the broader picture on AI and Wealth Management, that's been on our minds...as we thought about the Wealth space over the last couple of years...The winners in the Wealth space in respect of AI will be those who have scale and have data. When you think about 20 million customers that NatWest has, that's 200x the times of Evelyn. So the ability to use that scale and data [with] AI is a big accelerant and opportunity.

Secondly, what all the customer research and customer insight tells us, both independent and our own, is that the winning combination is going to be a combination of AI-driven digital Wealth advice, but also expertise through humans and people for those big financial decisions, the complex aspects of financial planning. So to me, you bring both together, you see AI really helping us get closer to our existing customers in the Wealth space, which is great, but also access new customers at relatively low marginal cost. So, net-net, we think AI will be an accelerant and a winner in terms of our Wealth aspirations.

Agentic

Agentic AI is also high up on the tech agenda at NatWest Group, according to Thwaite:

Our ambition is to become the leading bank delivering personalized customer propositions powered by the responsible deployment of agentic AI, so we are building out our capabilities across the bank. Last year, we set up an AI research office focused on improving customer experience and efficiency by accelerating the use of AI in fields such as multi biometrics, audio-visual conversational AI using proprietary small language models and ensuring algorithmic furnace as well as data safety.

This shift to a agentic AI marks a transition from simple chatbots to autonomous systems that can execute complex banking workflows on behalf of our customers. By prioritizing these capabilities, we can move beyond basic automation towards a simpler, data-driven experience that meets rapidly evolving customer expectations.

Many of the building blocks that will make this vision a reality will go live this year, he promises: 

This quarter, our customers will be able to ask questions about their recent spending in their own words on their app. And later this year, we will launch voice-to-voice conversations and more agentic fraud support. You can also expect to see a greater emphasis on the use of advanced data analytics to drive faster pricing, credit and asset enablement decisions. In addition, data analytics will help us manage risk dynamically, whilst optimizing risk-adjusted returns. 

All of this is coming at a cost, of course - Thwaite confirms that a further £100 million will be spent in 2026 on tech. What about the ROI on that? What levers are being pulled there to keep investors happy at a time when ‘show us the AI money’ is an all-too-familiar refrain? Thwaite argues:

It's a really broad range of levers, I would say. A key part of it is the kind of historic and current tech investment. That's driving a lot more digitization, automation. We continue to decommission a lot of applications. We've consolidated a lot of platforms. So that's really helping. We've also become a lot more efficient in how we do change...In effect, we can do more change at lower cost, which is great for the customer, but also great for the cost outlook. And we're also continuing to simplify the business more generally.

My take

Time will tell where that plays out...I do believe we've got this flywheel of kind of cost efficiencies going well...We’re very comfortable that this flywheel is heading in the right direction.

Onwards!