A guest editorial by Gemma Livermore, head of financial services international marketing at Seismic
Over the last few years, FinTech firms have sprinted toward artificial intelligence (AI), eager to automate everything from onboarding to customer support. However, it has recently become clear that these investments aren’t delivering the seamless experiences or cost savings many had initially expected.
As a result, many FinServ companies are coming to a key realisation: customers still prefer human support for certain interactions.
While investment in AI remains high, with financial institutions projected to increase AI budgets by 25% this year, these organisations are beginning to rethink their AI-first strategies as the tech fails to meet today’s consumer demands.
But what does this redirection mean for the wider financial services industry, and how can financial institutions continue to innovate with AI while still meeting rising expectations for trust, transparency, and regulatory accountability?
Where the cracks start to show: AI can’t do it alone
After going all-in on AI, Swedish fintech firm Klarna recently scaled back its end-to-end AI-driven customer service model.
Launched in 2023, Klarna’s AI-dependent strategy aimed to reduce operational costs by replacing up to 75% of its sales representatives with chatbots. While, in theory, this strategy promised efficiency, scalability, and stronger customer service – with the ability to respond in more than 35 languages and manage tasks like returns and payments – customer service quality greatly suffered, and satisfaction plummeted as a result.
After a year-long pause, Klarna has resumed hiring customer service representatives, now positioning quality human support as a competitive advantage, rather than a cost centre.
The lesson learned from all this? As acknowledged by Klarna’s CEO, this is a clear indication that customer expectations for human interaction still remain high, and that speed and efficiency do not necessarily equate to quality customer service. This is especially the case for the financial services industry, where trust and clarity are critical to brand loyalty and user confidence.
While, on its own, AI can’t support high-stake interactions with the same empathy and nuance as a human, there’s no question that it still has a beneficial role to play. So, how can it be applied in a way that facilitates a smarter and more valuable service?
Humans and AI: It takes two to tango
Klarna isn’t alone in this issue. Their reversal reflects a broader shift happening across the financial services industry. According to a Forvis Mazars’ report surveying UK FinServ C-suite executives, it was revealed that the industry’s AI adoption is predominantly driven by short-term experimentation rather than long-term strategic planning. This widening gap between ambition and execution is having a knock-on effect on implementation efforts, with 18% of institutions citing poor data quality as a top barrier in AI.
As institutions confront the limitations of AI-only strategies, many are now recognising that AI is not a silver bullet. While it is a powerful enabler, it’s not a replacement for human judgement, empathy, or contextual understanding. So, the real conversation is less about choosing between humans or machines, and more about finding the right balance between the two.
Forward‑looking FinServ institutions, like FIS Global and Royal London Asset Management, are embracing blended models as a way of enhancing human performance, automating the right tasks, and equipping customer-facing teams with the right content, tools, and insights at the right time.
This is the essence of modern enablement: aligning AI capabilities with human strengths to drive smarter, faster, and more personalised customer engagement. And it’s paying off, as industries implementing these enablement-led strategies are now seeing three times higher growth in revenue per worker.
But, with more advanced customer engagement comes more complex oversight – and that brings new regulatory considerations.
Staying compliant, customer-centric, and competitive
Any senior executive in financial services will likely tell you that regulatory demands are rising just as fast as customer expectations. Interestingly, Klarna’s decision to pull back on these AI-first initiatives arrives at a moment when buy now, pay later (BNPL) providers are about to face new regulatory scrutiny, especially around consumer protection and transparency in communications.
With the introduction of the UK’s upcoming BNPL regulations, institutions must now meet new obligations around affordability checks, fair treatment, disclosures, and complaint handling – all of which add complexity to customer-facing processes. In financial services, customer experience is tightly linked to compliance. Poorly governed automation can lead to poor customer satisfaction levels and increased risk of non-compliance. As such, FinServ firms can’t afford AI deployments that operate in silos from their compliance functions.
For institutions navigating these trust-heavy touchpoints, deploying AI can feel like walking on a tightrope. FinServ companies are expected to implement AI in a way that safeguards compliance while maintaining customer relationships and supporting long-term competitiveness. The key lies in integrating AI as part of a broader enablement strategy that provides customer-facing teams with contextualised and timely content to support informed conversations and decisions.
Implementing AI as an enabler
In regulated, relationship-driven industries like financial services, the margin for error is narrow. Gaining competitive advantage increasingly hinges on trust and personalised engagement.
To compete effectively, firms must reframe AI not as a front-facing fix, but as a behind-the-scenes enabler of smarter, more trusted interactions. This means steering away from simply slapping chatbots on the front line. Instead, companies should focus on building systems where every advisor, banker, or rep has AI quietly working in the background, surfacing compliant messaging, tailoring insights to the moment, and flagging risk before it escalates.
When grounded in enablement, AI earns its keep, helping firms meet today’s demands without compromising human touch.
Image: Marvin Meyer on Unsplash
Guest Editorial
This article was produced specially for Fintech Intel by an expert guest contributor.
