[ad_1]
The US author and philosopher, Elbert Hubbard, wrote: “One machine can do the work of 50 ordinary men. No machine can do the work of one extraordinary man.”
An interesting theory, but one that requires modern-day context. Hubbard died in 1915, when machine innovation primarily made manual tasks easier and more efficient. Machines certainly didn’t do your thinking for you.
Artificial intelligence (AI), as the cliché goes, has been something of a game changer. What AI is capable of now is extraordinary, but are we underestimating the speed at which, and extent to which, it will change this industry?
Bottleneck
Sikoia founder and CEO Alexis Rog argues that the real bottleneck to adoption of AI isn’t the capabilities of the technology — it’s the change in behaviour required to embrace it.
“The mortgage market is characterised by older target demographics and a heavy reliance on intermediaries like brokers, creating a Catch-22 between brokers and lenders,” he says.
AI is an enabling tool to enhance customer service, and not a replacement
“That said, I do believe we are underestimating the pace at which change will come in the market. Past adoption rates are not a true reflection of how quickly things are evolving now.”
Rog stresses that one of the biggest benefits of AI in today’s mortgage world is in improving efficiency and productivity, not replacing customer-facing advice.
“Traditionally manual tasks, like verifying income and employer, and assessing affordability from documents, can now be done in seconds,” he says.
“In the next five to 10 years, AI will help lenders and brokers handle more cases with the same resources, manage complex cases more cost-effectively and use customer data to offer better advice.”
At the recent Mortgage Business Expo 2024 in London, a seminar on AI adoption suggested that, over the next five years, there would be a sea change in the intermediary space and that, for 30%–40% of the mortgage market — essentially for vanilla products — AI would be the preferred option.
Data protection and data back-up systems must be watertight
Finova Brokers commercial director Matt Harrison says AI’s rapid evolution is already reshaping the mortgage industry, although its impact is more nuanced than some suggest.
“There are clear opportunities where AI can streamline tasks, like compliance and file checking, especially within the ‘vanilla’ mortgage segment,” he says. “But, for AI to fully step into complex advice roles akin to robo-advice, it would need to overcome significant regulatory and data management challenges.”
He adds: “Current digital solutions already effectively handle simpler cases, meaning sophisticated AI isn’t always necessary, and brokers remain essential for advising clients with more complex needs. Just because AI can do something doesn’t necessarily mean it should.
“We must also remember that the nature of financial services is underpinned with trust, which may be emulated via AI — but it has some way to go to replace human interaction.”
Generational factors
The issue of trust (or lack of it) is likely to have a say in how quickly AI is introduced and, indeed, accepted.
Recent research (November 2024) from Boon Brokers — ‘Artificial Intelligence in Mortgage Broking: Distrust from British Consumers’ — has found that most young people believe AI will fully replace human brokers within the next 10 years.
Proceed with caution before offering an AI service to older borrowers
Conversely, 88.3% and 89% of those aged 65-plus and 55–64 respectively would not trust technology to recommend a suitable product.
Boon Brokers managing director Gerard Boon says the discrepancy between the generations must be factored into advice provision because young people are far more likely than older generations to trial AI.
“For brokers who target first-time buyer business, they need to be aware of AI developments,” he says. “When the technology is proven and reliable, brokers should integrate AI into their service for first-time buyers to meet their evolving needs.
“However, for brokers who target older borrowers like lifetime mortgage clients, our survey indicates that offering a service with AI is likely to receive backlash and distrust. Proceed with caution before offering an AI service to older borrowers.”
Fignum CEO Andrew Lloyd agrees that AI buy-in does not lend itself to easy predictions.
“A decade ago, I worked as a management consultant. My team and I predicted that 30%–40% of mortgage business would by now be execution only. We were very wrong back then. It is often the case that we over-estimate the short-term impact but under-estimate the longer-term impact.”
The real threat is in over-reliance on AI, eroding the personal relationships that form the bedrock of our industry
Lloyd points out that, with mortgage loans, most people transact only every 20 years or so.
“It is a big decision and that means they want to talk to someone. The difference will be in the next 15 years, and it is to do with demographics — the TikTok generation don’t really like talking to people.”
Lloyd adds that Millennials and Generation Z are far more comfortable than older generations are with tech, and this will change how mortgage business is conducted.
Harrison does not wholly embrace the ‘generational’ theory.
“I believe AI adoption will depend more on how effective and useful it is, rather than on generational differences in tech comfort. I understand the competitive advantage firms might gain by integrating AI efficiently, but I see this evolution as no different from any other technological advancement.
My strong recommendation would be to look at framing some ‘acceptable use’ guidelines on AI for your firm
“Businesses that don’t leverage the latest tools will naturally fall behind, but ultimately I think AI will complement rather than replace the invaluable role of skilled brokers.”
Broker supremacy
Together director of intermediary sales Tanya Elmaz points out that AI is becoming increasingly integrated into business operations, and with this comes fear that it could replace ‘human’ jobs.
“Brokers and lenders alike are rightly cautious of what this new technology brings, and what the future holds,” she says.
Reflect on how much you trust the information in your hands and why you trust it
“However, AI is an enabling tool to enhance customer service, and not a replacement. It can be used to speed up processes or simplify onerous admin tasks for colleagues. It can help present a smoother and more efficient customer journey while navigating the often lengthy and complex process of borrowing.”
Elmaz adds: “There is a clear trend among mainstream lenders — which are more primarily driven by the volume of lending — that technology like AI will help to automate their processes entirely, delivering lending decisions with little human assessment. To stay abreast of this, brokers need to offer their customers solutions that AI simply won’t be able to deliver on, and one area of this is advice and counsel on specialist lending.”
Black & White Bridging director of lending Oli Bland agrees.
“For brokers, especially in the ‘vanilla’ segment, AI may open up new avenues to focus on specialist lending where their expertise and personal insight are truly needed,” he says.
“While AI will streamline standard cases, I believe it will allow brokers to concentrate on providing higher-value, tailored advice to clients.”
Fraud prevention
Importantly, AI can use machine-learning algorithms to detect potential fraud in mortgage applications.
Brokers and lenders alike are rightly cautious of what this new technology brings, and what the future holds
The argument is that AI can protect both the mortgage adviser and the customer, providing more security in the lending process. But does AI itself bring potential fraud risks?
Ohpen UK managing director Jerry Mulle is encouraged that AI can improve fraud detection but insists it also brings new challenges.
“The main issue is a lack of transparency and control, as these models often act as ‘black boxes’ that even their creators can’t fully explain. If fraudsters exploit these systems, it can be difficult to identify and fix.
“In contrast, rule-based systems offer transparency and easier fixes. However, they can also be exploited by fraudsters, who learn their rules.”
I believe AI adoption will depend more on how effective and useful it is, rather than on generational differences
Mulle agrees that, as AI advances in identifying fraud, so criminals will evolve their techniques.
“Where, on one side of the coin, there are benefits to be gained from AI being able to spot threats more quickly, on the other there’s the risk of scammers creating algorithms that can infiltrate the mortgage application process. So data protection and data back-up systems must be watertight.”
Group chief credit and money laundering reporting officer at OSB Group, Richard Wilson, agrees that firms need to be vigilant.
“Fraud with AI and Generative AI in essence relates to faking documents, faking voice and faking people [deep fake]. So, if you watch for these elements and use robust, trusted mechanisms, you are less likely to be a victim.
“Reflect on how much you trust the information in your hands and why you trust it. Do you think it is right or do you know it is right? If you know it is, why have you reached that assertion? Take a little time to stop and think.”
Wilson understands the concern regarding which data AI models can utilise, how it is exposed and where it can end up.
It is often the case that we over-estimate the short-term impact but under-estimate the longer-term impact
“If you are not careful, you inadvertently place data outside your controlled environment. My strong recommendation would be to look at framing some ‘acceptable use’ guidelines on AI for your firm, to avoid mistakes through lack of understanding.
“Common errors happen due to people not really understanding where data can end up when tapped into web-based engines or apps. Really build your understanding of what stays private and what can potentially move into the public domain.”
The way ahead
Looking ahead, Harrison says the sector must acknowledge that AI is increasingly accessible and user friendly — empowering non-technical users and opening up new possibilities for customer interactions, especially with tools such as large language models.
Brokers remain essential for advising clients with more complex needs. Just because AI can do something doesn’t necessarily mean it should
One evident barrier to introducing this technology is, of course, the cost, time and onboarding involved. As Rog highlights, many brokers and lenders are burdened with outdated systems, which hinder their ability to adapt quickly. Also, he suggests, a lack of general awareness of what is possible remains a key bottleneck.
Bland argues that, much as with previous tech advances, it’s all about finding ways to understand AI and then integrate it thoughtfully into the workplace.
“The real threat is in over-reliance on AI, eroding the personal relationships that form the bedrock of our industry. We must tread carefully lest we lose trust and human connection — those traits that clients rely on and value above all else.”
The mortgage market is characterised by older target demographics and a heavy reliance on intermediaries
Wilson sees things similarly.
“Technology will form part of the decision-making process and should enhance productivity.
“But it’s important to remember that the glue holding it all together is when the human element of underwriting really comes to the fore, as it can look at a much wider picture and assess the more subjective decision elements better, in my humble opinion, than a machine can.”
Perhaps Hubbard’s assertion from 1915 still applies?
This article featured in the December 2024/January 2025 edition of Mortgage Strategy.
If you would like to subscribe to the monthly print or digital magazine, please click here.
[ad_2]
Read more
