The UK rightly prides itself on having one of the most respected financial services sectors in the world but even strong markets benefit from looking outwards.
At Ohpen, we’ve been working strategically in both the Dutch and UK markets with our partners at Cognizant to reshape and redefine processes and at the centre of it all is AI.
AI is transforming work and life almost constantly but there are still many questions lenders have about how that transformation actually plays out in specific industries like mortgages.
Part of the problem is that AI has become a catch-all term. To understand its impact, we need to be more precise.
“AI isn’t just large language models like ChatGPT or Claude.”
AI isn’t just large language models like ChatGPT or Claude. It includes machine learning, natural language processing, computer vision, deep learning, robotics and expert systems. These technologies already underpin everything from virtual assistants to fraud detection and complex diagnostics.
The next step is agentic AI. Unlike traditional systems that analyse and recommend, agentic AI can also act autonomously within predefined policy and risk frameworks.
That distinction matters. Once AI is able not just to advise but to execute, the structure of an entire mortgage journey starts to look very different.
FRAGMENTED BY DESIGN
Today’s mortgage process is fragmented almost by design. Product design, sales, credit decisioning, underwriting, valuation, fraud and AML checks, legal work, operations, completion and servicing all sit in separate siloes.
Data is spread across multiple sources that don’t integrate easily. Some of those siloes belong to entirely different businesses in different industries. Frequently the underlying processes have barely evolved in decades.
This fragmentation is one reason open banking hasn’t delivered the step-change many hoped for.
Agentic AI, however, has the potential to do what open banking could not by turning the mortgage from a static product into a dynamic, adaptive service.
Longstanding concerns around regulatory risk and data governance are becoming less persuasive as controls improve and data becomes more reliable.
REIMAGINING THE PROCESS
The technology now exists to reimagine the process end to end. Instead of applying for products, customers authorise their data once and activate an AI agent.
That agent continuously gathers, validates and updates information from secure sources. In the Netherlands, this already includes PSD2, iDIN – a bank-backed digital ID – and links to tax, employer and asset data to create a live financial profile.
Within agreed parameters, agentic AI can assess affordability, flag anomalies and initiate follow-up actions automatically.
Real-time property valuations are pulled in via API connections to land registries and valuers, meaning borrowing capacity is calculated rather than estimated. What once relied on handovers and waiting periods becomes largely automated.
FUNDAMENTAL SHIFT
The result is a fundamental shift in experience. Lead times fall from weeks to minutes as friction disappears.
Mortgages stop being a standalone application process and become embedded in the home-buying journey itself.
As customers search for properties, financing options are already visible based on the data they’ve chosen to share. They can see borrowing capacity, indicative terms and monthly payments without committing to a full application.
“This level of automation doesn’t remove the need for advice.”
This level of automation doesn’t remove the need for advice. Instead, it makes clear where advice genuinely adds value.
Standard calculations and affordability checks fade into the background. What remains is interpretation, context and helping customers understand long-term risks, flexibility and future affordability. The adviser’s role shifts from product intermediary weighed down by administration to trusted guide at moments of real consequence.
ONGOING NOT EPISODIC
Agentic AI also enables advice to be ongoing rather than episodic. Decisions around fixed rate periods, repayments and flexibility can be revisited continuously throughout the mortgage term as life stage, career and risk profile evolve.
That requires an API-first infrastructure where origination and servicing work seamlessly together.
After completion, the relationship continues. AI can detect life events, propose proactive changes and link sustainability improvements directly to property value and mortgage terms. Prevention replaces reactive collections, aligning closely with the FCA’s consumer duty and its focus on vulnerability.
The UK may not yet have an equivalent to iDIN but we do have GOV.UK One Login, open banking APIs and a growing ecosystem of digital ID and KYC providers such as Onfido, Yoti and Trulioo.
The technology is already here. The real question is whether the mortgage of the future will be led by platforms, bank-owned AI agents or advisers willing to redefine where and how they add value.
