Artificial Intelligence is one of those topics that sparks every emotion imaginable, excitement, curiosity, fear, and sometimes a healthy dose of scepticism. Over the past few months, I’ve had conversations with brokers and industry peers about it, and one thing is clear: everyone is talking about AI, but not everyone feels the same way about it.
During a recent panel discussion at Mortgage Business Expo in London, I was asked to share my thoughts on how AI is changing the mortgage market and what that really means for brokers, lenders and technology providers. Within that discussion, I also spoke about the AI journey we’re on at Iress and what we’re learning along the way.
Here are some of those thoughts, which are also captured in a video interview you can watch below. And while my comments focus on mortgages, these themes apply equally across all areas of sourcing and financial advice.
Mixed feelings and why that’s a good thing
When I talk to brokers, the responses to AI are incredibly varied. Some worry it will disrupt their businesses, while others are genuinely intrigued and excited about its potential. Both reactions are completely understandable.
For me, there’s no doubt that AI, particularly at the pace it’s evolving, has the power to transform the mortgage market. The opportunity isn’t just about improving efficiency, but about enhancing the human experience for both the adviser and the client.
Technology should always serve people, not replace them. At Iress, we see it as our responsibility to help make that happen, and that starts with engaging openly with the people who use our tools every day, which we do through dedicated forums and daily interactions with clients. If we’re all part of shaping AI’s direction, it will reflect our shared values and goals.
How we’re bringing AI into Iress' technology
At Iress, we see two powerful ways to integrate AI into our sourcing technology:
- Internal acceleration – using AI to enhance how we develop software and deliver new capabilities faster.
- External adviser and consumer tools – creating intelligent “co-pilots” that support advisers in real time and help firms deliver a better client experience.
Firstly, AI presents a big shift for the software industry, bigger than the agile development movement. We’re entering a new frontier of software development. AI is not only helping developers build faster and better by working alongside them; it’s enabling the creation of applications that are more adaptive, intelligent, and interconnected in real-time. That’s incredibly exciting.
Developing safe, compliant and trustworthy AI for the mortgage industry will require collaboration between technology providers, lenders, distributors, and advisers.
Head of Product - Sourcing
Secondly, the AI opportunity for us lies in developing a specialist generative AI designed specifically for our industry. We know many of our clients are experimenting with AI tools like ChatGPT, Google Gemini, and Microsoft Copilot. There’s no denying these tools are impressive, especially when it comes to data analysis, but they’re generalists. As is to be expected, they can get things wrong, and they don’t (yet) have the nuanced understanding of financial products, regulation, and compliance that advisers need. These AI tools are also built on existing data and can contain inherent bias, which could result in poor and incorrect decision-making and recommendations.
Developing safe, compliant and trustworthy AI for the mortgage industry will require collaboration between technology providers, lenders, distributors, and advisers. But when we get it right, AI can truly become a co-pilot that enhances mortgage advice, not a substitute for it.
Safeguards and responsibility
With opportunity comes responsibility, and AI is no exception, especially based on the issues around accuracy and bias.
At Iress, we’ve established an AI Council to guide our strategy and ensure strong governance around how we develop and use AI. Every solution we bring to market goes through rigorous testing, is built on trusted data, and is launched with an abundance of caution.
For example, we would never launch an AI tool that sources live data directly from the internet. Accuracy, trust, and data integrity are non-negotiable.
What we’re learning along the way
It’s been an exhilarating learning journey. Every week brings new AI capabilities, and we’re continually exploring how they can add value to the advice process.
We’ve learned that not all AI is the same - and understanding which type of AI to use for each task is key.
- Generative AI (Gen AI) can create content, text, and code, but it requires precision in prompt engineering and robust fact-checking.
- Agentic AI takes it further - acting autonomously to complete multi-step tasks and learning from its actions along the way.
We’ve also learned that AI is only as good as the data and direction it’s given. Clear prompts, iteration, and human oversight are critical to avoid bias and misinformation.
AI: Threat or opportunity?
Both.
With so many new tools emerging, it’s easy to get caught in the trap of chasing the next shiny thing.
Head of Product - Sourcing
AI can absolutely accelerate how mortgage brokers operate, but only if implemented thoughtfully. It requires investment in upskilling, not just for your tech teams but across the entire organisation. And with so many new tools emerging, it’s easy to get caught in the trap of chasing the next shiny thing.
My advice is to learn about and evolve with AI tools, but stay focused:
- Identify the problems you want to solve
- Choose the right tools, and use them well.
- Remember trust and accuracy are everything, particularly in product and criteria sourcing.
There is a risk in getting it wrong. Not thoroughly testing, investigating and proving AI before use is a risky strategy that could harm trust and reputation.
Likewise, there is a threat in avoiding AI, because the industry and the technology aren’t staying still. If you don't work out a way for it to benefit you, a competitor will.
What AI shouldn’t do
Even AI agrees with me on this one. It shouldn’t take on low-volume, high-nuance tasks that require emotional intelligence, ethical judgment, or regulatory accountability.
AI should remain a co-pilot, not an autopilot, and currently is best deployed to handle the high-volume, low-nuance work extending to tasks like document processing and verification.
Revisiting the Robo-Advice debate
Naturally, conversations around AI bring back memories of the “robo-advice” debates from a few years ago. Back then, the industry grappled with whether technology could - or should - replace human advisers. And while there’s certainly a growing place for automated, digital advice in specific transactions and for certain client types, the reality is that both the technology and consumer appetite aren’t quite there yet.
At Iress, we see AI not as a replacement for advisers, but as an enabler.
Head of Product - Sourcing
Most people still prefer the reassurance of speaking with a human when making big life decisions, like buying a house. AI, for all its power, can’t yet replicate empathy or emotional intelligence - the ability to understand the motivations behind financial choices or the emotions that come with them.
That’s why, at Iress, we see AI not as a replacement for advisers, but as an enabler. We think of it as a co-pilot - augmenting the adviser’s expertise rather than supplanting it. The adviser remains firmly at the centre of the client relationship, with AI supporting them to deliver faster, more personalised, and more compliant outcomes.
Keeping it human
Balancing automation with a personal touch is one of the most interesting challenges in AI right now. Ironically, keeping things human often takes more time - from training AI to reviewing its outputs to refining its tone of voice.
Transparency will also be important. Just as we disclose when a meeting is being recorded, I believe we should be open when AI tools are part of the advice process. It’s simply another way of building trust.
What’s next? Touchless automation and integrated advice
Over the next 12 to 24 months, I expect to see “touchless automation” become a reality for many non-complex mortgage applications.
Technologies like Intelligent Document Processing (IDP) and Open Banking will make it possible to process applications end-to-end with minimal human intervention - freeing advisers to focus on complex, value-added interactions.
And here’s where it gets even more exciting: as AI understands a client’s mortgage liability and verified income, it can also identify their protection gap in real time.
That means protection advice can be seamlessly integrated into the mortgage journey - helping advisers deliver more holistic, responsible outcomes for clients.
Final thoughts
AI is here to stay, and it’s moving fast - not just in the mortgage world, but across all of financial services and the world at large. While we might not be able to fully comprehend the full impact AI will have in the long term, its near future success in our industry will depend on how responsibly and collaboratively we approach it.
For me, the focus isn’t on the fear of machines replacing humans. It’s about using technology to empower and serve people, to build trust, and to deliver better outcomes for every client.
Let's take this opportunity to help shape the future impact of AI on Mortgages and financial services together.
