“If I had asked people what they wanted, they would have said faster horses.” This line is usually credited to Henry Ford. The Henry Ford museum, however says it “has never been satisfactorily traced” to him (The Henry Ford). None the less it survives because it captures a real temptation: when a new technology arrives, we use it to speed up the thing we already do.

Professional services are at that point with AI. The question I keep coming back to is not whether firms should adopt it. They already are. It is whether bolting AI onto the systems and processes we have will deliver much, or whether the bigger gains need us to rethink how the firm’s technology, and its work, are put together.

The adoption numbers are no longer in doubt. In the Bank of England and FCA’s 2024 survey, 75% of UK financial services firms were already using AI, with a further 10% planning to within three years (Bank of England and FCA, 2024). The lang cat’s 2026 State of the Advice Nation research found AI use among UK advice firms rose from 29% to 60% in a year, and to 88% at firms managing more than £1bn (the lang cat, 2026, cited in Professional Adviser, 2026). Lawyers are moving just as quickly: a LexisNexis survey found AI use in UK legal work rose from 46% to 61% between January and August 2025 (LexisNexis UK, 2025).

AI use in UK advice firms doubled in a yearShare of advice firms using AI
UK advice firms, 2025
29%
UK advice firms, 2026
60%
Firms managing £1bn+2026
88%

Source: the lang cat (2026), State of the Advice Nation, cited in Professional Adviser (2026).

The impact numbers are less comfortable. BCG’s 2024 study of 1,000 executives found 74% of companies had yet to show tangible value from AI (BCG, 2024). MIT’s NANDA initiative reported in 2025 that around 95% of organisations it studied saw no measurable profit-and-loss impact from generative AI, and pointed to tools that integrate poorly with existing workflows (MIT NANDA, 2025). That report has been criticised for its method, and I wouldn’t lean on the exact figure. But the direction is consistent with Gartner’s 2024 prediction that at least 30% of generative AI projects would be abandoned after proof of concept, because of poor data, weak risk controls, rising costs or unclear value (Gartner, 2024).

Look at how the technology is actually used and the gap makes sense. The Bank of England and FCA found that 62% of AI use cases were rated low materiality and only 2% involved fully autonomous decisions. Almost half of firms (46%) said they had only a “partial understanding” of the AI they use (Bank of England and FCA, 2024). In advice firms, the main uses are admin, meeting notes, data handling and report drafting (the lang cat, 2026, cited in Professional Adviser, 2026). These are useful. They are also, mostly, faster horses: the same steps, done quicker.

Economic history has been here before. In 1987 Robert Solow wrote that “you can see the computer age everywhere but in the productivity statistics” (Solow, 1987). With the invention of electricity factories first replaced their steam engines with electric dynamos and kept everything else: the same line shafts, the same layout, the same flow of work. The large gains came only when factories were redesigned around a motor on each machine, and in the US that took decades (David, 1990).

Erik Brynjolfsson and colleagues have since formalised the pattern as a “productivity J-curve”. A general-purpose technology needs heavy investment in things that don’t show up on a balance sheet, such as new processes, skills and data. Measured productivity dips before it rises (Brynjolfsson, Rock and Syverson, 2021). In other words, the technology is rarely the bottleneck. The organisation around it is.

The current evidence points the same way. BCG found that the companies getting value from AI put about 10% of their effort into algorithms, 20% into technology and data, and 70% into people and processes (BCG, 2024). McKinsey’s 2025 State of AI survey found that redesigning workflows is one of the strongest contributors to business impact, and that the highest performers were nearly three times as likely to have fundamentally redesigned individual workflows (McKinsey & Company, 2025).

AI leaders put 70% of their effort into peopleHow companies getting value from AI split their resources
People and processes
70%
Technology and data
20%
Algorithms
10%

Source: BCG (2024), Where’s the value in AI?

Financial advice has a particular version of this problem. A typical firm runs its client record in a back-office system, its money on one or more platforms, its planning in separate cashflow software, and its communication through email and diaries. NextWealth’s adviser research has found lack of integration to be the biggest pain point in back-office technology, and back office has long sat at the bottom of advisers’ satisfaction tables (NextWealth, 2025, cited in Professional Adviser, 2025).

AI makes this problem more visible. A model that drafts a suitability report is only as good as the fact-find it reads. A meeting-notes tool saves time, but if its output is re-keyed into three other systems, much of that saving is lost.

Microsoft’s chief executive, Satya Nadella, made a related point in late 2024. He described business software as databases with business logic on top, and argued that the logic will increasingly move into AI agents working across those systems (Nadella, 2024). If he is right, the value shifts from the individual application to the layer that connects data and work across the firm. My view is that this is exactly where the most exciting area of financial advice technology is right now. The evidence favours rethinking the work and the data first, and replacing systems selectively. Practically, that means four things.

Start with the workflow, not the tool: map how a piece of advice actually moves through the firm, from the first meeting to the annual review, and ask where the handoffs and re-keying are. Fix the data layer before adding intelligence on top: one reliable view of the client, available to every system that needs it. Keep humans accountable for decisions, and design the oversight in rather than adding it later. And measure outcomes, such as time returned to client work or files that pass checks first time, not the number of AI tools in use.

Ford may never have said it, but the warning still holds. Faster horses are a real improvement. They just aren’t the thing that changes the industry.

Maxwell & Partners is a financial services technology provider working with financial advice firms to improve efficiency and expand the quantum of customers advisors can serve. The company is currently in stealth mode.

References