New research puts AI almost level with human advisers as a comfortable place to ask financial questions. The bigger challenge is what happens to the business model when clients need fewer answers from you.
For thirty years, I have watched a pattern repeat across financial services.
As consumers gain visibility, understanding and the means to act for themselves, intermediaries come under pressure to justify their place.
Charges become easier to see. Products become easier to compare. Information becomes easier to obtain. Tasks that once required a professional become things people can do themselves.
That does not explain every merger, branch closure or adviser exit. Regulation, costs and commercial decisions matter too. But it raises a question the industry cannot keep answering with reassurance:
When a client can do more, what are they still paying you to do?
AI brings that question into the consulting room.
A one-point lead should prompt a bigger conversation
Research commissioned by money app Plum, reported by Financial Planning Today on 16 October 2026, asked consumers whom they would feel most comfortable approaching with a financial question.
A traditional financial adviser came first, at 19%. An AI assistant followed at 18%.
Banks and search engines each attracted 16%, followed by family at 15%.
This was an online survey of 2,022 UK adults, conducted by Research Without Barriers between 27 and 30 July 2026.
It did not establish that AI provides equally good advice. Nor did it measure willingness to pay, actual usage or the preferences of existing advice clients. Without the full statistical analysis, we should not attach significance to a one-percentage-point lead.
But the near parity is a signal worth taking seriously.
For a sizeable group of respondents, a machine is already a comfortable starting point for a money question.
The first barrier may be embarrassment
The research also found that 53% had avoided asking someone a money question because they feared appearing financially inexperienced. The same proportion had pretended to understand a financial term or conversation.
Think about that from the client’s side of the desk.
You are paying someone for expertise. You do not understand their explanation. Asking again might expose how little you know.
So you nod.
An AI assistant offers a different experience. Ask the basic question. Ask it again. Request an example. Challenge the explanation. Nobody looks impatient.
That does not guarantee accuracy. It can, however, make learning easier to begin.
And it reveals something uncomfortable about traditional relationships:
A client with no questions may be well informed. Or they may be too embarrassed to ask.
Silence has never been a reliable measure of understanding.
What happens when the questions really do disappear?
Now consider a different client.
They arrive having used AI to understand the terminology, compare charges, organise their records and explore possible courses of action. They have checked sources and identified what remains uncertain.
Some questions they once brought to an adviser have already been resolved.
Perhaps they need twenty minutes of focused help where previously they needed several meetings.
That is progress for the client.
For a business whose fees depend partly on being the place where answers live, it creates pressure.
A few fewer questions will not necessarily reduce the value of a relationship. Better preparation can lead to deeper conversations. Some people will continue to value delegation, continuity and professional accountability.
But if the client needs less work, perceives less benefit or can obtain a satisfactory result elsewhere, their willingness to pay may change.
We cannot celebrate efficiency inside the firm while assuming efficiency on the client’s side has no commercial consequences.
“We provide the human touch” needs a second sentence
Empathy matters. So do judgement, reassurance and an understanding of someone’s life.
But naming those qualities does not establish the value of a particular service or fee.
What did the conversation help the person understand? Which assumption needed challenging? What consequence had they overlooked? What could they decide or do afterwards?
Consider someone weighing up retirement.
They may already understand their spending, pensions and broad options. Their remaining difficulty could be a disagreement with their partner, uncertainty about purpose or a specific technical issue that needs qualified expertise.
There is worthwhile work here.
The fee should be explained through that work and its benefit. A general claim that humans understand humans cannot justify every charge attached to the relationship.
Nor does valuable planning automatically justify the cost of the investment channel bundled with it.
Each part of the bill must earn its place.
Build a service that benefits from capable clients
This is the direction we have pursued at the Academy of Life Planning since 2012: helping people become more capable in their own lives.
AI expands what becomes possible.
A consumer-owned planning system can help someone maintain their information, revisit assumptions and prepare decisions between conversations. Human support can then concentrate on the problems that warrant it.
Within our ecosystem, Total Wealth Plans provides the AI operating system, while Total Wealth Planners provide human support when needed.
The commercial task is to make that support clear, useful and proportionate.
Sometimes that means education. Sometimes coordination, a difficult conversation or help turning an intention into action. Sometimes it means recognising that a decision requires a suitably qualified specialist.
Success should include a client needing less help with things they have learned to handle.
A model that requires continuing confusion has a problem when understanding becomes easier to obtain.
Who benefits from the client’s AI?
Plum has its own commercial interest: it offers an AI planning tool linked to financial products. Its research deserves attention, alongside recognition of that context.
The wider question applies to every provider, including us.
Does the system help people compare openly, question its suggestions and leave with usable information? Or does growing confidence mainly make it easier to move them into the provider’s products?
An AI interface alone tells us little about whose interests the underlying business model serves.
The test is whether the person gains practical control.
The fee question is already here
This study does not prove that advisers are becoming redundant. It does suggest that the industry can no longer assume people will naturally turn to a human first.
As clients become more capable, some will ask fewer basic questions. Others will bring better, harder ones. Some will want less support. Others will recognise exactly where expertise is worth paying for.
Build for that client.
Help them own their information. Welcome their scrutiny. Make the contribution visible. Let the fee reflect the work.
When your client already has the answers, what can they achieve with you that they could not achieve as well without you?
That is the fee conversation worth preparing for.
Source: Financial Planning Today, “19% prefer human advisers but AI close behind”, 16 October 2026, reporting Plum-commissioned research by Research Without Barriers. Survey details and findings above are drawn from the supplied article. Independent coverage: Finextra, “Lack of financial confidence driving consumers to AI assistants”.
