It’s been a few weeks since we’ve really dug into the “next big thing” here. As I mentioned in my last post, after a bit of a hiatus, it’s exciting to be back observing the emergence and merging of banking, finance, and technology. And frankly, there has never been a more disruptive moment to jump back in.
You can’t have a conversation about tech today without tripping over “Generative AI.” For most of the public, and even for many in our industry, the immediate mental leap is to the customer-service chatbot.
We’ve all seen the demos: a slick, conversational interface that can answer “What’s my balance?” or “Explain a mortgage.” This is useful, certainly. It’s an evolution of the clunky, rule-based bots we’ve tolerated for years.
But this is the tip of the iceberg. And as is so often the case in finance, the most profound changes aren’t the ones on the customer-facing billboard. The real revolution is happening in the middle and back office—in the complex, data-heavy, and regulation-drenched engine room of the bank.
Years ago, we were all talking about “Big Data” and its potential (You can scroll down to my post on this from a few years ago). Generative AI is the engine that finally unlocks it. It’s moving us from analytics (what happened) to automation and augmentation (what to do next).
If you want to see where GenAI is really changing the game, in financial services at present, look past the chatbot and focus on these three areas.
1. Supercharging Compliance and RegTech
For any modern financial institution, compliance isn’t just a department; it’s a foundational cost centre and a critical risk. The sheer volume of new and updated regulations is staggering.
This is where Large Language Models (LLMs) shine.
- The Problem: A bank’s compliance team has to read, interpret, and implement hundreds of pages of dense regulatory text from multiple jurisdictions. They must then monitor internal communications (emails, chats) and transactions for potential breaches of rules like KYC (Know Your Customer) and AML (Anti-Money Laundering).
- The GenAI Solution: Instead of rules-based systems that flag keywords (“guarantee,” “promise”), GenAI models can understand context and intent. They can read a 300-page regulatory update and instantly generate a summarised “to-do” list of policy changes. They can scan communications and flag a conversation that feels collusive or evasive, even if it doesn’t use specific trigger words. This isn’t about replacing the compliance officer; it’s about giving them a team of a thousand junior analysts, allowing them to focus on high-level judgement.
2. The New Frontier of Underwriting and Fraud
We’ve used AI in risk modelling for years, but it’s largely been a game of structured data—credit scores, transaction history, income. This approach is powerful but blind to the world of unstructured data.
- The Problem: A traditional underwriting model for a small business loan knows the company’s cash flow, but it doesn’t know about the new competitor that just opened across the street, the negative sentiment in local news, or the supply chain risk mentioned in an industry report.
- The GenAI Solution: GenAI can read, understand, and summarise all of it. An underwriter can now get a rich, contextual summary in seconds: “Applicant is a 5-year-old cafe with stable cash flow. However, recent local news articles show a 30% decline in foot traffic for its neighbourhood, and two new competitors have opened within a mile in the last six months.” The human makes the final call, but now they’re armed with a complete picture. The same applies to fraud, where models can now spot suspicious narratives and connections between seemingly unrelated accounts, not just isolated, anomalous transactions.
3. Personalising Wealth Management at Scale
True, hyper-personalised financial advice has historically been the exclusive domain of high-net-worth individuals. It required a human adviser to spend hours analysing a client’s portfolio against their personal goals and the entire universe of market data.
- The Problem: It’s impossible to offer this level of service to the mass market. There just aren’t enough hours in the day.
- The GenAI Solution: We’re now seeing the rise of the “AI Co-pilot” for financial advisers. This tool can analyse a client’s complete financial picture, ingest their stated goals (e.g., “retire in 10 years,” “save for a house”), and monitor real-time market news. It can then draft personalised insights for the adviser: “Based on today’s interest rate hike and Client Smith’s risk profile, here are three high-yield bond ETFs to discuss with them. Here is a summary of the pros and cons for each, and a draft email to start the conversation.”
This isn’t a robo-adviser replacing the human; it’s a “robo-analyst” that empowers the human to be a true strategist for every single client, not just the wealthiest.
The Mutual Angle: A Different Set of Stakes
Now, for the UK’s building society sector, the “emergence and merging” of this tech has a unique flavour. A global bank can invest billions in GenAI to chase efficiency. A building society, owned by its members, has a different calculation: how to innovate without losing its identity.
Their brand is built on trust, community, and a personal touch—assets that are easily damaged by poorly implemented tech. The temptation to just “install a chatbot” to keep up with the digital-first banks is huge, but it’s a trap. Research shows members, particularly the long-standing ones, are wary of AI and value human interaction.
For the mutual sector, the real impact of GenAI isn’t about replacing that human touch; it’s about supercharging it.
- Smarter Mortgage Underwriting: This is the core business. Instead of just structured data, a GenAI-assisted underwriter can analyse and strucure local council planning documents, property surveys, and even company accounts in seconds. This doesn’t replace the human underwriter’s decision; it gives them a “superpower,” allowing them to make a more informed, nuanced judgement on complex applications like for a self-employed person or a Ltd Company that a big bank’s algorithm might simply reject. It’s an efficiency tool that doubles as a competitive advantage.
- Compliance on a Mutual’s Budget: Building societies face the same crushing regulatory burden as the high-street giants (Consumer Duty, AML, fraud) but with a fraction of the compliance staff. This is where GenAI is a massive force multiplier. It can automate the drafting of regulatory reports, scan for fraud, and process member complaints (a key part of Consumer Duty) with a speed and accuracy that a small team simply cannot match. It helps them “punch above their weight” in risk management.
- Delivering on the “Member-Centric” Promise: The “AI Co-pilot” concept is even more critical here. It’s not about “wealth management” for the few, but about “financial well-being” for all members. Imagine a branch colleague, empowered by an AI tool, who can instantly see a member’s full history and get a prompt: “This member’s fixed-rate savings account is maturing. Based on their goals and our current products, here are two options to discuss.” It’s technology used to foster a more meaningful human conversation.
For building societies, GenAI is a tool that allows them to solve their digital dilemma: it provides the back-office efficiency to compete, while simultaneously empowering their staff to double down on the very thing that makes them different.
The theme of this blog has always been the “merging” of finance and tech. For the first time, this isn’t just about digital channels or faster databases. It’s about embedding intelligence into the very core of what a financial institution does: manage risk, ensure compliance, and provide advice.
The chatbot is just the interface. The real story is the engine. And it’s an engine that’s in the process of being completely rebuilt—for everyone, from the global banks to the local building society. The only question now is how quickly firms can adapt their culture and processes to leverage it.
Ready to discuss? I’d love to hear your thoughts on this, or what you’d like me to deep-dive into next. Let me know in the comments or drop me a note!
MN
Finovation 