UK Regulator Urges Fintechs to Prioritise Customer Needs Over AI Hype

The Financial Conduct Authority released its latest assessment of artificial intelligence use in financial services, spotlighting strong practices alongside critical weaknesses. The watchdog stressed that firms must define customer needs before developing AI products, keep effective human oversight in place, and build operational resilience into their systems.

Sophie Njagi, chief executive of UK-regulated electronic money institution EQWIRE, responded by urging fintech companies to embed those principles into every stage of product development. Her message: commercial excitement around AI risks creating a chasm between what dazzles in a demo and what actually functions when deployed at scale.

Products can look impressive in a demonstration but struggle when customers start using them at scale. The test is whether they solve a meaningful problem, make the customer's experience simpler and have the operations behind them to deliver reliably.

While the FCA did not name specific firms, its review revealed a recurring pattern: financial services businesses are rolling out AI capabilities without sufficient understanding of the customer challenges they intend to solve, and oversight structures have not evolved to match the technology being introduced.

Operational resilience emerged as a major area where AI implementations fall short. The concept has been embedded in formal rules by the FCA and Prudential Regulation Authority for several years, yet many firms' AI deployments lag behind those expectations.

The findings carry particular significance given the regulator's evolving stance. The FCA has signalled support for AI innovation but made clear that the Consumer Duty, which took full effect in 2023, obligates firms to demonstrate positive outcomes for retail customers. That transforms customer-need identification from a design preference into a compliance requirement before allocating engineering resources.

Njagi outlined five principles she believes underpin technology products built for durability:

• Prioritise genuinely important customer problems over fashionable technology trends
• Invest in operational foundations rather than surface-layer design
• Manage complexity internally so customers experience simplicity
• Build teams with genuine ownership of outcomes
• Maintain strategic flexibility to adapt as regulation, technology and customer behaviour evolve

While these principles are not groundbreaking, the FCA context adds a compliance layer that pure product advice typically lacks. For fintech founders and product leaders, the regulator's review signals that AI roadmaps will face scrutiny based on customer outcomes, not solely on technical risk management.

The FCA's review arrives during a period of rapid AI adoption across retail banking, payments and lending. Both established banks and challenger firms are integrating large-language-model tooling into customer-facing channels and back-office workflows.

The regulator's primary concern appears to be firms adopting AI capabilities that perform impressively in controlled settings but whose real-world performance, edge-case handling and failure modes remain poorly understood by those deploying them.

For smaller fintechs and electronic money institutions operating under tighter capital and compliance budgets, the operational resilience mandate carries extra weight. Constructing the oversight infrastructure regulators expect around AI systems introduces costs that must be justified by the customer benefits delivered. Njagi's perspective, as CEO of a regulated EMI, frames the discipline demanded by regulation and the discipline required by sound product development as fundamentally aligned.