Commercial chief says the industry needs to rethink AI – not as a threat, but as a route to doing more
Artificial intelligence is reshaping what is possible inside a mortgage lender, but the debate around whether it replaces people is missing the point entirely, according to a senior figure in the buy-to-let sector.
Steve Cox (pictured top), chief commercial officer at Fleet Mortgages, a specialist buy-to-let lender owned by Starling Bank, believes the industry has become too preoccupied with AI as a threat to jobs, when the more useful frame is what the technology makes possible operationally. For Fleet, the answer is straightforward – more lending, processed faster, without proportional growth in headcount.
"What we will look at is more use of technology and AI to create efficiencies," Cox told Mortgage Introducer. "You can view AI as a job killer, but actually it's a capacity giver."
Where AI fits – and where it doesn't
Cox is precise about what he expects AI to do, and what he does not. At Fleet, where the majority of lending involves limited company structures and complex assets, the underwriting process demands a level of human judgement that he does not see technology replicating in any meaningful timeframe.
"Do I see AI underwriting a limited company buy-to-let on a complex asset with complex company structures? No, I don't. That's not going to happen. People need to do that."
The use cases he does see as genuinely viable are more targeted – reading valuation reports, cross-referencing data at Companies House, processing documentation that would otherwise require an underwriter's time without adding to the quality of the decision.
"Can I see AI being able to read a valuation report that comes in and tell the underwriter the answer without the underwriter having to look at it? Yes, absolutely. But what that's doing is creating efficiency. We have capacity to do more lending, do it faster, simpler and quicker, still get to the right decision, but without having to expand the headcount."
That view aligns with the broader direction of travel in UK financial services. A joint survey by the Bank of England and the Financial Conduct Authority (FCA) found 75% of UK financial firms are currently using AI, with another 10% planning adoption within three years – a sharp increase from 58% in 2022. Across the sector, the priority areas identified by institutions – optimising internal processes, improving operational efficiency, and enhancing customer support – map closely onto the efficiencies Cox describes.
Figures from the Office for National Statistics show the proportion of UK businesses with at least 10 employees using AI rose from around 12% in late 2023 to 35% in June this year. But analysts have noted the range of AI tools being actively deployed has lagged behind headline adoption rates. For specialist lenders, the question is not whether to use AI, but where it genuinely adds value without compromising the rigour that complex underwriting demands.
The human element in specialist lending
Cox extends the same logic to the broker market. Brokers, he said, are already using AI to improve the efficiency of their processes, but the technology is not replacing the advice relationship, and he does not expect it to.
"Brokers are using AI to create efficiencies in their process, with regulated advice. But it isn't replacing the person."
The distinction he draws is between commodity and complexity. For straightforward cases, a vanilla residential product transfer, a simple remortgage at low loan-to-value on a well-known property type, automation is plausible and, in his view, probably inevitable. But for specialist lending, on either the lender or the adviser side, human expertise remains the deciding factor.
That position is consistent with what brokers themselves have been saying. Research published in June found one in three brokers are now comfortable with greater use of AI or automation in parts of the mortgage journey, with document verification and administration identified as the areas of highest appetite. Scepticism persists, however, with 20% of brokers saying AI would not have a meaningful positive impact on the mortgage process.
Steady growth, sharper tools
Fleet's broader ambitions for the next year or two are deliberately measured. Cox describes a strategy of steady growth rather than aggressive expansion – a stance that has defined the lender's approach since its acquisition by Starling. Within that, investment in AI and technology represents one of the clearest levers available to grow lending capacity without a corresponding increase in cost.
The FCA launched a review in January into how advanced AI may affect consumers, retail financial markets and regulators over the rest of the decade – a signal that the regulatory framework around AI in financial services is in active development. For lenders already embedding the technology into their operations, understanding where human oversight remains non-negotiable will be as important as the efficiency gains themselves.
Cox is unequivocal on where the line sits. "I can't see AI replacing advisors. I just can't see it. People buy people."
Want to be regularly updated with mortgage news and features? Get exclusive interviews, breaking news, and industry events in your inbox – subscribe to our FREE daily newsletter. You can also follow us on Facebook, X (formerly Twitter), and LinkedIn.
