Would hours of study and years of experience be consigned to the dustbin of history as borrowers shunned human intermediaries and looked online for guidance on their mortgages? Would face-to-face advice soon be a thing of the past?
Happily, none of this came to pass: 84 per cent of mortgage business in the UK was arranged through an intermediary in 2022, and the Intermediary Mortgage Lenders Association believes that figure is set to rise to 90 per cent.
A pretty convincing testament to the enduring indispensability of sentient mortgage advisers.
But artificial intelligence (AI) is accelerating at a breakneck speed – so much so, that even heavily-invested figures such as Elon Musk earlier this year signed an open letter calling for a pause in development, warning that ‘AI tools pose a profound risk to humanity and society’.
Cynics believe he just wants the breathing space to catch up with the competition. But 1,000 other tech luminaries also put their names to the document.
Should mortgage advisers be concerned?
Let’s look at recent example of AI’s advancement: the chatbot, which allows people to talk with AI directly.
The current market leading chatbot, Microsoft’s ChatGPT, doesn’t just answer questions but can summarise text, write essays in various styles and has even passed a range of degree-level medical exams. It is scarily impressive, particularly when given a tight brief.
But, certainly as yet, it is not capable of offering personalised mortgage advice.
For the purposes of this article, I asked ChatGPT whether I should fix my mortgage for two or five years. It responded (in about five seconds) with five recommendations, which boiled down to: consider the interest rate outlook; think about my own financial stability; consider my future plans as a homeowner; evaluate my risk tolerance and review any fees and penalties on a mortgage deal.
All of those are fair points, but unhelpfully generic as advice goes.
Just as well that its closing lines were: “Ultimately, it’s essential to consider these factors alongside your personal financial goals and preferences. It can also be helpful to consult with a mortgage adviser who can provide tailored advice based on your specific circumstances and the current market conditions.” For now, at least, AI seems to be on the side of the intermediary.
When I asked ChatGPT to identify its own limitations, the bot listed them clearly: lack of common sense and human experience; inability to understand context; bias based on the data it was trained on; limited world knowledge; lack of emotional intelligence; inability to reason.
No AI can (currently) understand nuance or feel empathy. That list explains why AI won’t be replacing human mortgage advisers any time soon.
But what AI can do is automate routine tasks like data inputting and document processing and analyse data from a range of sources to arm advisers with insights into market trends and customer behaviour. Machine learning algorithms can detect fraud, enhancing the security and the integrity of adviser firms. All of which can not only improve the quality of customer service intermediaries can offer, but free up a huge amount of time which can be focused on giving the tailored, individual advice many borrowers need.
The importance of offering the highest possible standard of advice and ensuring the best customer outcomes is only growing with the advent of Consumer Duty.
The more we can leverage AI’s capabilities to complement and support adviser interactions, the better.
