Earlier this year, the launch of an AI tax planning tool briefly wiped significant value off the share prices of several wealth managers. This reaction reflects genuine concern about AI disrupting the sector, but I think it’s overstated.
Simple, standalone tools may be vulnerable to rapid change, but complex, regulated platforms are an entirely different proposition.
Wealth managers are dealing with millions of customers and investments running to many billions of pounds. At that scale, any new technology has to work within the controls already in place. Every firm has its own policies and risk frameworks, and AI needs to fit around them rather than replace them.
The Financial Conduct Authority has made it clear that ignoring AI isn’t an option. Its chief data, information and intelligence officer Jessica Rusu recently warned that firms which don’t adopt AI quickly risk falling behind criminals already using the technology against consumers.
AI adoption has to be approached with the same consideration for client needs as any other change that affects how they are served.
The industry needs to spend less time worrying about what AI might do in theory and more time looking at applications that work under normal operating conditions.
That means how it handles exceptions, how it responds to incomplete data and what happens when the process doesn’t follow the expected path.
AI that only performs well in a controlled setting is not ready to be deployed. It needs to operate consistently within existing systems, including when unusual cases come up.
AI is unlikely to change the fundamental nature of how advice is delivered. The biggest impact for advisers will probably be on the speed and reliability of the administration behind every client interaction.
Advisers face regular delays when dealing with providers and platforms and for AI to make a measurable difference, firms must be able to trust it.
The frustration advisers experience with platforms and providers often comes down to how much of the work required is still manual. The problem is that traditional rules-based automation requires standardisation, but in wealth management different firms have unique ways of doing things.
Pension transfer requests are a good example. They need multiple checks and validations, and each firm handles the steps differently. AI, though, can follow a firm’s own procedures consistently, without changing how it operates or forcing everything into a one-size-fits-all workflow.
In this way, AI will reduce the manual work and costs embedded in wealth managers’ operational processes, with advisers and clients benefiting from the efficiency gains, including faster onboarding, smoother transfers and fewer delays.
Much of the concern around AI focuses on autonomous systems that can make open-ended decisions and take actions without human input. While these worries are understandable, they don’t reflect how AI needs to work in a highly regulated environment where firms have to keep control over decisions and how risks are managed.
In regulated firms, AI should be used for tightly defined tasks, with people stepping in when anything falls outside the agreed rules. It should stick to approved actions and escalate exceptions, not make free-form decisions. Wealth managers need to be able to understand what action has been taken and why. Solutions that can’t provide that evidence won’t progress beyond the pilot phase.
At the moment, too much of the discussion is about AI disrupting technology. The focus should be on how it can improve day-to-day operations and reduce risk and delays.
Investment platforms and wealth managers have come a long way in automating routine processes in the past 15 years, but human intervention is still required in too many situations, adding risk, cost and delay.
The responsibility wealth managers carry for customers’ long-term financial security leaves no room for shortcuts.
Rob DeDominicis is GBST’s Chief Executive Officer, leading the group’s strategic direction and driving continued growth across its global business. This article first appeared in FT Adviser on 09/07/2026. You can read it here.
