The Rise Of The AI-Empowered Customer: Financial Services' Blind Spot

This article was previously published on Forbes.

The financial industry is one popular app away from a dramatic shift in power between institutions and consumers, and most don’t even see it coming. 

Seventy-five percent of U.K. financial firms now use AI, up dramatically from just two years ago. Insurance companies lead the charge at 95%, with international banks close behind at 94%. Billions are flowing into algorithms that promise faster underwriting, better fraud detection and lower operating costs. 

While firms race to embed AI into underwriting, fraud detection and cost-reduction programs, they’re overlooking the far bigger disruption: not how financial institutions use AI, but how consumers will use it against them. 

Firms have spent years perfecting AI to manage customers. What they haven’t prepared for is customers using AI to manage them. 

This isn't a hypothetical future. Early warning signs are already here. 

Complaint volumes have surged in the U.K., driven by claims management companies using templates and automation to industrialize complaints. This wave was eventually containable because intermediaries were regulated and charged fees for their cases. But when customers bring their own AI, those guardrails disappear. Consumers can raise a complaint free of charge, and if successful, they are typically awarded a payout. There is currently no regulation that prevents millions of customers, each equipped with their own AI advocate, from raising queries, disputes and complaints. 

Picture 2027: A new £9.99-per-month personal AI finance assistant hits the market. It monitors accounts 24/7, automatically budgets and flags any inconsistencies in fees, interest rates or product terms. 

It spots a £42.50 overcharge hidden in a mortgage payment three months earlier. Within seconds, it generates a complaint that references contract clauses, regulatory obligations and Ombudsman precedents—written with legal-like precision and impossible to dismiss with a standard reply. 

Now imagine tens of thousands of these landing across the industry every day. 

The nightmare isn’t that every complaint is correct; it’s that they will sound correct. Each will require specialist review, data validation and careful responses. Under the Consumer Duty, firms must thoroughly investigate and respond appropriately. The cost of disproving even invalid AI-generated complaints could escalate from a few pounds to hundreds. Multiply this at scale, and customer service infrastructure collapses. 

Banks will need armies of specialists to respond to consumers’ AI agents. 

This shift creates three crushing forces: 

1. Rising Costs: Handling complex AI-generated complaints within regulatory timeframes will strain systems and require heavy investment in people and technology. These costs will inevitably be passed on to consumers. 

2. Eroding Trust: If consumers must choose between a message from their bank and a message from their personal AI, most will trust the AI. Consumer technology has already conditioned people to expect clarity, accuracy and responsiveness that many financial institutions struggle to match. 

3. Regulatory Exposure: Consumer Duty makes no distinction between complaints written by humans and those written by AI agents. Firms must demonstrate fair treatment and proper investigation, even when facing mass-produced, technically sophisticated queries. 

To get into the AI race, many financial firms are trying to bolt AI capabilities onto legacy architectures that weren’t designed for real-time, context-rich, AI-to-AI interactions. Although financial firms regularly use AI for automated decisioning or scoring, this is very different from operational AI, which involves processing unstructured, adaptive communication from multiple sources. Most existing platforms aren’t designed to support these use cases. 

Legacy systems can’t provide instant, auditable, contextually aware responses. Considering that some firms have spent decades trying to modernize legacy tech stacks without success, the truth is unavoidable: Old architectures won’t support the emerging AI-driven environment. 

The winners will be those who centralize and operationalize their data on modern, event-driven platforms with AI built natively at their core, not bolted on around the edges. 

So, what should you do? 

1. Build AI on unified, real-time context. Agentic AI means customers' personal assistants will interact with your systems constantly, requesting breakdowns, updates, clarifications and personalized explanations in real time. Delivering accurate responses requires a single, consolidated source of truth that instantly brings together all relevant data and processes. Fragmented systems cannot meet this bar and will lead directly to Consumer Duty risks. 

2. Embrace radical transparency. Products, fees and terms must be defensible and straightforward. If AI cannot find contradictions or confusion, neither can regulators or customers. Clear contracts and accessible data aren’t just good service—they’re essential protection. 

3. Engage regulators early. The industry needs shared definitions for frivolous AI-generated complaints, standards for automated dispute handling and frameworks for AI-to-AI interaction. Regulators will expect firms to manage this new complexity, and collaboration will be essential. 

You should consider forming or joining industry consortia to work with the FCA, Financial Ombudsman Service and HM Treasury on developing frameworks for AI-to-AI communication within financial services. The regulatory system wasn't designed for the volume or sophistication of complaints on the way. 

Financial services have long relied on information asymmetry; institutions knew more about products and terms than customers did. This advantage is disappearing. 

AI is putting granular product knowledge, contract analysis and behavioral monitoring directly into consumers’ hands. The question isn’t whether customers will challenge every fee, discrepancy and clause. They will. 

The real question is whether your firm will be ready. 

The age of the AI-empowered customer has already begun. Most of the industry is still looking the other way.