3 ways AI in banking and finance is transforming the digital customer experience

AI in banking and finance is nothing new, but there’s nothing ‘business as usual’ about the extent to which the financial services (FinServ) industry has embraced AI in recent years.

Research from Lloyds Banking Group shows two-thirds of U.K. financial institutions invested in AI in 2024, almost double the number from the previous year. 

And in its 2025 Financial Institutions Sentiment Survey, Lloyds reports 91% of U.K. financial institutions now view AI as an opportunity rather than a threat—with more than half (51%) planning to increase their AI investment in the year ahead.

This surge is largely fueled by recent advances in AI technology, particularly the growth of generative AI tools. Generative AI in banking is big business, with management consulting firm McKinsey predicting it could deliver value to the industry equal to an additional $200 – $340 billion annually.

AI has a wide range of applications in banking, but this article will focus on one that’s of critical importance to banks: the optimization of digital customer experience (CX).

Customer expectations around digital experiences are a constantly moving target. What amazes customers today will be the expected minimum tomorrow. However good a business’s digital CX is, it can (and needs to) get better, and AI will help them make that happen. 

Let’s take a look at 3 AI applications in finance and how they’re helping banks create more engaging, seamless, and retention-driving digital journeys. 

1. AI chatbots offer answers now—and soon, AI agents will guide customers and businesses

It’s well-documented that AI chatbots benefit both customers and customer service agents. 

According to a study cited by McKinsey, implementing generative AI at a company with 5,000 customer service agents

• Increased issue resolution by 14% an hour

• Reduced the time spent handling an issue by 9% 

• Reduced agent attrition and requests to speak to a manager by 25%

Impressive stuff. But this is only the start of what chatbots can achieve. 

Today’s chatbots are great at answering simple customer inquiries, such as requests for account information. In the very near future, however, they’re set to evolve and offer customers personalized advice, functioning as virtual advisors.

For some fascinating insights into the future of chatbot technology in banking, check out this talk from Tasneem Bhamji, Digital Engagement Director at Lloyds Banking Group, at Contentsquare’s recent CX Circle events. (It’s also embedded below.)

Tasneem speaks about how Lloyds Banking Group is using AI to improve its customer experiences right now—and how the Group is currently working towards a more sophisticated, personalized, and intimate AI-powered service. 

The future of customer-facing AI agents with Lloyds Banking Group

Currently, Lloyds Banking Group has an AI chatbot embedded in its app that can answer relatively simple questions. 

But the Group knows customer expectations are set to rise dramatically in the not-too-distant future. As a result, their chatbot will have to evolve into an AI agent, capable of providing more complex advice. 

“When it comes to AI, customers don’t just want a one-dimensional answer anymore,” says Tasneem. “They want to be guided through things. In terms of how we use AI in our customer experiences, for us, this means showing up differently than we do today.”

When you’re buying a complex financial product, whether that’s something really big and emotional like buying a house or if you’re extending your overdraft, you need to feel that you’re being guided in a trustworthy way. Our AI guide has to earn the right to help customers by earning their trust, by making it clear how and why it’s arrived at the guidance it's giving them.

Tasneem Bhamji

Digital Engagement Director

There are 3 core pillars guiding Lloyds Banking Group in its approach to creating an AI guide that customers will feel confident taking advice from. 

1. Transparency: it needs to tell customers exactly why it’s suggesting something

2. Consent: it shouldn’t tell customers what to do—customers rightly don’t like this, especially when it comes to the emotional and personal matter of money

3. Empathy: it needs to make clear that what it’s suggesting or recommending is for the customer’s benefit, not the bank’s

2. Predictive analytics will enable true personalization

Personalization is about leveraging customer data to serve experiences that cater to their specific interests, desires, and needs—and nothing leverages data like AI.

AI analytics lets you analyse user behaviors at scale, identifying patterns and trends in your data that empower you to make informed predictions about future behaviors. This means you can anticipate customers’ needs and interests before they’ve expressed them (and potentially even become aware of them).

Customers can be provided with suggestions (‘Would you like to open a savings account?’) and recommendations (‘Check out our beginner’s guide to investing in stocks and shares’) with the confidence that they’ll respond positively to them. 

Plus, issues can be anticipated. For example, AI can identify app users who are in danger of churning and proactively engaging them to ensure they stick around. 

In the future, AI’s ability to analyze data and make predictions based on this analysis will help banks personalize their web and app experiences to unprecedented levels, anticipating customers' needs before they even arise. 

Using AI tools for analytics to know the customer will be what separates the leaders from the pack in digital banking. Tasneem Bhamji agrees.

Everyone is talking about how fast AI can serve an answer. But it isn’t speed that will differentiate customer experiences in the future—it's trust. You build trust by knowing the customer. That means not just giving them the fastest answer, but giving them the most useful answer based on your knowledge of them, and by talking to them in their language, not like they’re a segment.

Tasneem Bhamji

Digital Engagement Director

3. Automated workflows speed up time to insight—and automated analytics will kick things up a gear

Banks are using AI to automate time-consuming processes, whether that’s running financial reports, underwriting loans, or collecting payments.

Workflow automation directly improves the customer experience by speeding up services and cutting down on human errors that can sometimes occur, causing customer frustration. It also indirectly benefits CX by freeing up time for banks to focus on improving customer service and digital user journeys. Even better news: many of the processes that go into monitoring, analyzing, and optimizing the customer experience are also now being automated.

As with all things AI, this is really only the beginning as far as automation goes. 

Today’s AI needs to be prompted and given pre-defined tasks. Tomorrow’s AI (while still being guided by goals set for it) will make its own decisions, plan actions, adapt to changing circumstances, and learn from its own mistakes. 

It’ll have its own agency—hence the name it's been given: agentic AI.

In other words, much of digital experience analytics will itself be automated—and the impact on your team’s productivity and the quality of your customer experiences can (for now) only be imagined. 

Meet the future of AI in banking and finance CX, today

According to Contentsquare’s “How to Beat the 2025 Benchmarks for Financial Services” ebook, many finserv experiences are in need of some serious optimization. 

For example:

• 56% of all finServ sessions end abruptly with a bounce, meaning customers leave after viewing just one page

• 32.7% of all finServ sessions are marred by some form of customer frustration 

This is particularly bad news given traffic to finServ sites dropped -4.3% year-over-year in 2024, while the cost per visit jumped +7%. 

To beat bounces, fight frustration, and take full advantage of every increasingly pricey visit, your teams need granular visibility into your behavioural data. 

And that means all your teams—not just the analysts. 

When only data analysts can perform data analysis, this creates bottlenecks. These slow down time to insight and lumber your analysts with low-level requests when they could be innovating. 

Fortunately, Contentsquare's AI-powered experience intelligence platform is here to accelerate and democratize data analytics. 

Sense (Contentsquare’s AI) automates numerous analytics processes, creating effortless analytics workflows that radically accelerate time to insight. These include

• Session Replay Summaries: see what happens during one or multiple sessions in one click, highlighting key insights, issues, learnings, and even recommended next steps

• Voice-of-Customer (VoC) survey summaries: see an at-a-glance overview of customer sentiment with key recurring positive and negative trends highlighted

• Real-time alerts: set up alerts when Sense detects unusual variations in your behavioral data

This all helps make analytics faster and more accessible, but what Contentsquare offers that really democratizes data analytics is AI assistance, courtesy of the built-in generative agent Chat with Sense.

Anyone in your business—whether in marketing, product, sales, or C-suite—can ask Chat with Sense a question about your digital experience data and get an answer almost immediately.

That answer comes complete with charts, summary of key findings, and (if relevant) insights into the measurable business impact of issues and opportunities. Plus, you can ask follow-up questions (perhaps picking one from the list of suggested next questions that Chat with Sense provides for you).

Sense is evolving fast and is expected to become agentic in the future. It’ll independently analyze your behavioral data and proactively advise on what you need to do to optimize experiences. 

Check out our Buyer's Guide to Contentsquare's AI to find out how Sense makes digital experience analytics faster, more impactful, and accessible to all (including the not-so-tech-savvy).

Alternatively, see for yourself—and book a demo today. (You don’t need ChatGPT to tell you it’s a good idea.)