Nathan Reilly, Chief Customer Officer, for Twenty7tec, explains that whilst artificial intelligence is now more central than ever in the sector, it cannot replace the human-to-human element that is central to effective mortgage advice.
Artificial intelligence is no longer theoretical for the mortgage and property investment sector. It is already influencing research and data analytics tools, client communications and operational workflows. The real question is not whether to use AI, but where it genuinely adds value.
In practice, AI works best when it removes friction from repeatable tasks rather than replacing conscious human judgment. Mortgage firms handle high volumes of documentation, case notes and client correspondence. AI can summarise calls, draft follow-up emails, organise meeting notes and flag missing information. That reduces administrative time and allows advisers to focus on client relationships and case strategy. Much like having the right software reduces friction in your activity day-to-day, AI reduces the need to use the brain power on smaller tasks when instead, it can be used to converse with customers and create lasting relationships.
In research and sourcing, AI can analyse large datasets, identify product shifts and highlight relevant criteria changes. But it should not replace suitability assessment. Advice involves nuance, context and accountability, especially in a world now where simple cases are rarely a consistent occurrence.
AI delivers real value when it strengthens human expertise rather than substitutes it. The biggest gains today are in efficiency, monitoring product changes, summarising interactions and analysing documentation. But advice and underwriting involve judgment and responsibility. Those cannot and should not be delegated to an algorithm, especially as the nature of the job involves people’s homes, and probably the largest purchase they will ever make.
Compliance and Consumer Duty remain central. AI outputs must be transparent and auditable. Firms need clear governance, defined accountability and oversight processes before scaling adoption. Bias monitoring and escalation routes are not optional extras. Removing the friction of being able to be compliant means it is actually easier and faster to follow the rules.
There is also a risk of overstatement. Not every process needs automation. AI is most effective when applied to specific, measurable challenges such as reducing tedious work, improving document handling or monitoring product volatility. Over time, roles will evolve, and advisers may spend less time gathering information and more time interpreting it. Skills such as critical thinking, communication and risk awareness will become even more important rather than “switching our brains off” in favour of letting AI handle everything for us.
The opportunity for the sector is not full automation. It is intelligent augmentation accompanied by human interaction and advice. In a regulated industry built on trust, that balance matters.
