Artificial intelligence (AI) is rapidly changing the world. Even in school classrooms, teachers are sifting through essays, trying to work out whether they were written by students or by ChatGPT!
The mortgage industry is no exception. AI-powered mortgage lending platforms are already being used by some lenders in the UK, and their adoption is expected to grow in the coming years.
You really do get out what you put in
One of the key advantages of AI in the mortgage market is the ability to streamline and automate complex processes. AI-powered algorithms can quickly analyse vast amounts of data and make decisions in seconds, reducing processing times and increasing efficiency.
Patterns and correlations
AI can also help to improve the accuracy of lending decisions by identifying patterns and correlations that would be difficult for humans to spot. By analysing a multitude of variables and historical data, this should lead to potentially more reliable risk assessments and reduce the chance of defaults for lenders.
Furthermore, AI is being used to enhance the customer experience with chatbots and virtual assistants that are intelligent beyond anything we’ve seen.
It’s less about the ‘magic’ and more about investment in the right data, the model architecture, and the meticulous evaluation and optimisation of the results
They provide personalised responses 24 hours a day, removing the need to sit ‘on hold’ in a phone queue, waiting to speak to a human and hoping the call isn’t accidentally disconnected.
British Telecom recently revealed plans to significantly reduce its workforce as part of an effort to cut costs, which includes an estimated 10,000 roles being replaced by AI, according to its chief executive, Philip Jansen. Goldman Sachs, meanwhile, predicts that AI could replace the equivalent of 300 million full-time jobs worldwide in the future.
Balanced view
Taking a balanced view, for centuries technology has been a big factor in changing the labour market. Manual jobs have been replaced by machines the world over, dating back to the invention of the printing press and beyond. The term Luddite is still used today as a whimsical slur for those who are opposed to technology, and of course new jobs will be created with the growth of AI.
It’s vital that we train our AI using trusted sources of data to cultivate its growth
Job losses will grab the headlines along with the claim that ‘robots will take over the world’, but that perhaps isn’t the biggest or most realistic threat.
There is the obvious risk that a lack of human judgement in the underwriting process may impede borrowers with unique circumstances that can’t be accurately evaluated by AI alone.
Algorithms are only as good as the data they’re trained on. If that data is biased, it may lead to discriminatory practices against certain borrowers, such as those with low credit scores or a limited financial history.
Inaccurate information
However, more worrying at present is the ‘information’ available to the end customer that can be found on the general AI platforms where they may be tempted to seek ‘mortgage advice’.
The team at Knowledge Bank have tested some of these platforms with regard to the information around mortgage criteria and affordability, and have been staggered by the inaccuracies.
Taking a balanced view, for centuries technology has been a big factor in changing the labour market
Although this made for an enlightening (and at times entertaining!) exercise, the concern is that the general public won’t be able to validate the accuracy of the answers as easily as we can, which will lead to problems and inevitable disappointments.
I firmly believe there is a place for AI within the mortgage industry and it can be used as a force for good, but it has to be harnessed and nurtured responsibly. The likes of ChatGPT and Google’s Bard, which are readily available to the public, are not trained on proprietary databases and don’t have access to information behind paywalls or login screens.
It’s an exciting time for our industry and we have a real opportunity to build technology that will not only generate efficiencies but also create new opportunities for lending. But it’s vital that we train our AI using trusted sources of data to cultivate its growth, as this is one scenario where you really do get out what you put in.
One of the key advantages of AI in the mortgage market is the ability to streamline and automate complex processes
It’s less about the ‘magic’ and more about investment in the right data, the model architecture, and the meticulous evaluation and optimisation of the results.
These will be the key factors to AI’s success within the mortgage industry. Get it right and the possibilities are infinite.
Nicola Firth is founder and chief executive of Knowledge Bank
This article featured in the July/August 2023 edition of MS.
If you would like to subscribe to the monthly print or digital magazine, please click here.
