Lenders have the timing advantage at remortgage, but AI is starting to level the playing field for brokers
Around 1.8 million fixed-rate mortgages are due to expire this year, and lenders are already positioned to keep most of that business.
According to UK Finance data, internal product transfers, where borrowers renew with their existing lender rather than switching, reached £256 billion last year, against just £71 billion in external remortgaging.
For mortgage brokers, that gap represents a persistent and largely unresolved revenue problem. Annmarie Baylan, a former broker, told Mortgage Introducer AI is the only tool capable of closing it.
Baylan is the founder of Go Automation AI Ltd, a WhatsApp-based communication platform for Financial Conduct Authority (FCA)-regulated mortgage brokers. Her argument is not that brokers are failing their clients, it is that the structural odds are stacked against them, and AI is the only tool capable of rebalancing them.
The lender's advantage
The problem, as Baylan sees it, is one of timing and design. Lenders know exactly when a borrower's fixed rate expires and act on that information with precision. Brokers typically do not have equivalent systems, and even when they attempt to follow up manually, the volume of clients with different end dates and different lender schedules makes consistent outreach close to impossible alongside running an active advice business.
"The lender makes it really easy," Baylan said. "They know exactly the date the mortgage expires, they're going to contact you and send you a button, and you click on the button – it's done. They're pulling the wool over their eyes a little bit really."
The broker's position is made harder still by the variation in how lenders handle rate release timing. "Halifax release that date at four months, Platform at six months, Barclays at three months," Baylan said. "They are all different." A broker who contacts a Halifax customer at five months before their expiry date and finds they have no additional needs has no rate to work with and faces the task of remembering to follow up again at exactly the right moment. More often than not, that call does not happen.
What can AI realistically do for brokers?
Baylan's contention is AI can address this not by replacing the broker's role but by solving the logistical problem that prevents the broker from playing it. In her view, the adviser's value is in the human conversation that converts. The relationship built over time, the judgement applied to complex circumstances, and the advice that a lender product transfer cannot replicate. What AI can do is ensure the broker is in front of the client at the right time, with the right information, when that conversation needs to happen.
"All AI is doing is assisting the broker to get in front of the customer at the right time with the right information so that they can compete with the current lender," she said. "It can't have that human relationship, and it never will have, in my opinion."
The timing question is, for Baylan, the crux of what she built her platform to solve. Clients with additional needs – further borrowing, a planned move – need to be reached at six months before expiry, when there is still time to do the necessary work. Those with no additional needs should be booked closer to the date their specific lender releases retention rates, so the adviser walks into the meeting with a real number to compare against the market. Without that precision, a broker can do everything right and still arrive too early to act.
Consumer Duty and the compliance case
Beyond the commercial argument, Baylan points to Consumer Duty as a driver for more structured client contact throughout the fixed-rate term. The FCA's ongoing requirements oblige firms to demonstrate that clients were meaningfully informed, and their circumstances understood before advice was given. For many brokers, the honest answer is this level of consistent contact simply does not happen.
"Consumer Duty says you should keep in touch with the customer throughout the term," Baylan said. "A lot of people don't, just because it's difficult to get organised sometimes."
The compliance dimension has added urgency in the current environment, with how brokers can embrace AI without losing the human edge an increasingly live question for firms of all sizes. Baylan is clear the answer is not to hand client relationships over to automation, it is to use automation to protect the conditions in which those relationships can happen. She noted that centralised, timestamped communication records also address a vulnerability many firms overlook – a broker who uses a personal WhatsApp account to contact clients takes that conversation history with them if they leave.
For those still unconvinced automation is compatible with quality advice, a recent study on whether AI can really replace mortgage brokers found that only 12% of buyers say they would fully trust a standalone AI platform – a figure that underlines where the value of human advice still sits.
The broader question for the profession is not about threat or replacement. It is about whether brokers have the tools to compete with institutions that have spent years engineering frictionless switching at the precise moment it matters most.
"It's about embracing AI as a partnership," Baylan said. "It's about it working for you. It's not going to take over."
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.
