Better banking needs humans and AI
Capgemini ·
Discover how banks can use AI to improve customer experience, unify fragmented journeys, reduce service costs, and combine automation with human expertise. The post Better banking needs humans and AI appeared first on Capgemini .
For banking leaders, the question is no longer whether AI can improve customer experience (CX). It’s where AI should take the lead, where human judgment must remain visible, and how both can work hand in hand.
You may recall that at one point merely providing digital access to customers was enough for a bank to earn its CX stripes. But today, customers place far greater value on the quality of that access.
Since banking interactions often carry real financial consequences for the customer, weak CX isn’t just an inconvenience. It compounds into slower resolutions, repeat contacts, and escalations, each adding direct cost to serve in addition to the risk of reputational damage.
We found that for 88% of banking executives CX is central to their growth strategy. But is delivering it consistently such a straightforward task?
40% of banking executives cite “fragmented journeys across multiple channels” as the greatest barrier to effective CX today, with 58% expecting it to worsen over the next three years.
Fragmentation typically shows up at the handoffs customers notice most, such as when switching from a mobile app to a branch or call center mid-task or re-submitting identity documents already provided elsewhere in the institution.
Each handoff adds friction for the customer and rework for the bank. And because many of these journeys touch regulated processes such as KYC (know your customer) and account changes, the cost of getting them wrong can create compliance issues and leave teams tracing records and correcting documentation.
But disconnected journeys aren’t the only challenge. Banks may also be designing experiences around assumptions that overestimate how happy customers really are. Our research shows that 96% of executives believe customers are willing to recommend their products and services based on past experiences. But in reality, not even half ( 48% ) of customers actually do.
Left unaddressed, this gap shows up as weaker referral-driven acquisition, missed cross-sell opportunities, and customers who quietly shop competitors at their next major financial decision, exactly the outcomes banks believe their CX is already delivering.
Now that AI shows up everywhere, from search engines and shopping apps to workplace software, customers are becoming more comfortable with AI agents handling routine banking tasks, especially when those tasks reduce effort without increasing perceived risk.
For example, 33% of banking consumers say they may soon let AI agents do things like fill in loan or credit-card applications for them. That creates a practical opening for banks to improve onboarding, servicing, and everyday interactions.
These figures point to where banks are concentrating automation first: high-volume, low-complexity servicing, such as balance inquiries, card activation, payment status, and application intake. This frees staff for interactions where judgment and reassurance are needed.
As AI adoption accelerates, banking customers will continue to rely on human expertise for complex decisions. Our survey reveals that almost three-quarters (72%) of customers rank human support among their top three preferred interaction channels. At the same time, 59% rank AI-integrated bots among their top three, showing a clear preference for hybrid experiences, not an either-or choice.
But this hybrid preference is not evenly distributed. Customers are far more willing to let AI handle routine, low-stakes tasks, such as checking a balance or tracking a payment, than they are to navigate a loan denial, a fraud claim, or a hardship request without a person on the other end.
Banks that route by moment rather than by channel – matching the mode of interaction to the emotional weight of the task – can help customers feel more comfortable using AI for the tasks it’s best suited to.
For one financial services company, we used intelligent routing to determine when an interaction should be handled by AI and when it needed a live agent. This helped reduce annual call volumes by a third, from 300,000 to 200,000, while improving a CX that had been falling short.
We’re also developing an agentic desktop that adapts to the caller, transaction, and reason for contact, giving employees the right information at the right time. By making interactions easier for employees to manage, the company can provide faster, more relevant experiences for customers. A happy employee makes a happy customer.
The objective isn’t to remove people. It’s to apply human judgment where it makes the most sense and let AI handle all the routine interactions that typically create administrative hassle.
To deliver a consistent experience, banks, card issuers, lenders, and wealth managers need one connected CX strategy rather than a collection of channel or product initiatives. That strategy should focus on applying AI to the journeys that matter most to customers and drive the greatest cost – from onboarding, KYC, everyday banking, and payments to card and lending servicing, disputes and fraud, collections, and wealth and advisor servicing.
A unified approach also creates a consistent model for where AI leads and where people add the most value. Routine interactions and transactions can move to conversational and agentic AI, while frontline employees and advisors use AI copilots to support more complex financial decisions, exceptions, fraud, complaints, retention, and advice.
Underpinning all of this is a trusted data and context foundation that connects mobile and web, contact centers, branches, advisors, and operations. When insight and interaction history travel with the customer, each channel becomes part of the same journey, so customers do not have to reauthenticate, repeat information, or re-explain their needs at every stage.
Explore our latest CX report to understand how expectations are changing as AI becomes more embedded in banking. Then talk to us to learn what these findings mean for your specific organization and CX strategy.
We would be happy to arrange a customized presentation just for your team.
Research and Insights Report World Wealth Report 2026 Author Chandramouli Venkatesan Vice President – Portfolio Development Lead – Digital Front Office Transformations | Banking and Capital Markets Chandra leads the Front Office transformation portfolio (marketing, sales and customer service) and serves banking and capital markets clients. He focuses his work on customer experience and helping financial institutions transform marketing, sales and customer service into more customer-centric organizations with an emphasis on experience strategy design, technology and data. Chandra has deep experience driving CX transformation for retail banks, payments companies, wealth management and capital markets firms. FAQs Where should banks use AI, and where should human expertise remain? Banks can use AI for routine, high-volume interactions such as balance inquiries, card activation, payment status and application intake. Human expertise remains particularly important for more complex or higher-stakes situations, including loan denials, fraud claims and hardship requests. The opportunity is therefore not to replace people, but to combine AI efficiency with human judgment where it adds the most value.
Banks can create a more connected experience by ensuring customer data, context and interaction history move across mobile and web channels, contact centers, branches, advisors and operations. This reduces situations where customers have to reauthenticate, repeat information or explain their needs again when moving between channels.
Customers are becoming more comfortable allowing AI agents to handle routine banking activities when doing so reduces effort without adding perceived risk. The research cited in the blog finds that 33% of banking consumers may soon allow AI agents to complete activities such as filling in loan or credit-card applications on their behalf.
Customers show demand for both human and AI-supported interactions. The research finds that 72% of banking customers rank human support among their top three preferred interaction channels, while 59% rank AI-integrated bots among their top three. This points toward a hybrid experience in which AI handles suitable routine interactions while people remain available for situations requiring judgment and reassurance.
Banks can move beyond separate channel or product initiatives by building a connected CX strategy spanning journeys such as onboarding, KYC, payments, lending, disputes and fraud, collections, and wealth servicing. Routine transactions can move toward conversational and agentic AI, while frontline employees and advisors can use AI copilots to support more complex decisions and customer needs.
The post Better banking needs humans and AI appeared first on Capgemini .