Five years ago, banks were already using AI to automate routine tasks, from fraud detection and customer service chatbots to back-office processing. The benefits were clear: faster decisions, greater efficiency, and reduced manual effort. While the technology has evolved dramatically, the industry’s core objective remains largely the same.
Today, this landscape has scaled dramatically. AI adoption in the financial sector is near universal, with only 2% of firms not using the technology. As AI-powered processes are estimated to cut costs by 30% to 45%, it’s becoming increasingly clear that focusing on efficiency alone can be limiting.
When AI is layered onto fragmented systems and poor data quality, expected returns often fail to materialize, sometimes at the expense of customer satisfaction. Banks need to broaden their success metrics to include customer experience, retention, and long-term value creation to build a sustainable future.
The chatbot trap
Modernization is at the forefront of every financial organization’s AI strategy. With challenger banks and fintechs driving fierce competition, improving operational efficiency and reducing costs are top modernization priorities for 82% of firms. AI’s promise of significant cost savings makes efficiency-led deployments attractive.
The reality, however, is more complex.
Financial institutions increasingly use AI as a gateway to customer service, with chat interfaces becoming the default route for everything from routine inquiries to complex issue resolution.
From a financial viewpoint, the model is attractive, offering immediate efficiency gains while letting banks redirect resources elsewhere. However, when banks treat AI ROI solely as operational cost-reduction, they risk overlooking the broader value the technology can unlock.
For “lower value” retail customers, AI-driven chat support is the primary, and sometimes only, customer service option, with human support reserved for high-net-worth clients.
In my experience, the trade-off is increasingly difficult to ignore:
- Lower servicing costs
- Reduced access to human support
- Rising customer dissatisfaction
- Higher customer turnover
When customers cannot easily escalate issues to a human representative, AI can become a source of friction rather than convenience. As resolution times rise and frustration grows, trust erodes. Banks must ask whether short-term cost savings are worth the long-term impact on customer loyalty.
Beyond faster processes
In the race to adopt AI and deliver better customer value, many organizations have focused on deploying new tools rather than addressing the underlying conditions for success. According to BCG, only 5% of organizations report generating value from AI at scale.
Considering the scale of investment poured into these initiatives, the challenge is clearly not a lack of ambition. Yet financial institutions keep increasing spending. The challenge is not a lack of investment, but a failure to rethink how work gets done.
Too many institutions now view AI as a technology project rather than a business transformation initiative. They move quickly to integrate AI into existing operations without first redesigning the fragmented workflows beneath them.
As a result, AI often accelerates inefficient processes instead of eliminating them. Even the most sophisticated models will struggle to deliver meaningful value when the data they rely on is incomplete or spread across siloed systems.
This helps to explain why so many AI initiatives fail to scale. The greatest returns stem from rethinking how information flows through the organization, how decisions are made and how employees interact with customers.
Rather than viewing AI as a standalone productivity tool, financial institutions should treat it as an enabler of broader, long-term transformation across data, systems, and workflows.
The loyalty question
Banks often measure AI success through immediate efficiency gains. Yet the greatest returns may come from stronger customer satisfaction, better service, and deeper product engagement. So, why are banks still evaluating AI ROI primarily through an operational lens?
The risk is that efficiency becomes the objective rather than the enabler. Faster, cheaper processes only create value if they enhance customer experience and deepen relationships.
As AI adoption matures, banks should ask a simple question: Is it helping customers stay, engage, and grow their relationship with the bank, or merely reducing workloads?
The most forward-looking institutions will integrate AI not simply to cut costs, but as a strategic driver to make smarter decisions, deliver stronger customer outcomes, and unlock new sustainable growth opportunities.
Mike Stawchansky is chief technology officer at Finastra.