Visibly demonstrated human expertise and judgement matter more when clients use AI.

Could AI create better clients?

Far from being an existential threat to advice, it’s just possible that artificial intelligence (AI) could be the best thing that ever happened to it.

Through an adviser lens, the efficiency benefits of AI are already significant, helping create a welcome downward pressure on the cost to serve and lifting practice capability in areas including communication and record keeping.

But perhaps the biggest – industry-reshaping – benefits are those viewed through a consumer lens. By making it easier to source detailed knowledge about financial products and constantly changing tax, super and Centrelink rules, AI will not only reinforce the need for human expertise to navigate the complexity, it will also mean clients come into advice engagements equipped with more base knowledge, and able to ask better questions.

Russell Investments’ 2026 Value of an Adviser research^[1] suggests that shift is already under way.

Their study found that 57 per cent of non-advised investors use AI to support financial decisions¹, most commonly to research financial topics and explain complex concepts. The proportion saying AI will make them more likely to seek professional advice is greater than the proportion saying it will make them less likely (26% v 19%), and only 4 per cent cite AI tools such as ChatGPT as a barrier to seeking advice^[2].

As one adviser told Russell’s researchers:

“I love that my clients are coming in and asking me the right questions. We can get to the deeper questions, the things that actually matter, because they are arriving with the foundational knowledge already in place^[3].”

Increasingly, adviser value is sitting in the relational aspects of advice, a fact already recognised by advisers themselves. In Russell’s study, 47 per cent say their main focus is managing client relationships, against 5 per cent who say selecting investments.

Clients can now easily find explanations and explore options for themselves. They want the adviser to apply their expert judgement to that information. It may not be coincidental that another finding in Russell’s study relates to the drivers of client satisfaction. In 2026, the biggest driver – by a considerable margin – was expertise, with its index score rising to 139, from 118 a year earlier, ahead of the next highest driver, tax-savvy planning, at 101. As information gets easier to find, clients are increasingly rewarding the judgement applied to it.

This article explores how advisers can make their expertise visible: how AI is changing the advice conversation, how judgement tests a plausible-sounding answer, and how advisers can use AI in their own work while retaining responsibility for the advice. A case study shows the thinking in practice. Unless otherwise noted, all findings quoted are from the 2026 edition of the Russell Value of an Adviser report.

What is expertise, anyway?

Russell describes a broader meaning of expertise, beyond technical competence alone: “the ability to apply judgement when the answers are not obvious”.

Russell also describes expertise as both technical and emotional, built over years, and visible in how an adviser understands the way different parts of a client’s financial position interact. It names three roles through which advisers apply their expertise throughout the client relationship: the Guide, who helps clients weigh competing priorities; the Guru, who turns information into advice; and the Gladiator, who protects clients’ interests when the stakes are high.

Access to information may be getting easier, but applying judgement to the information is not, and for advisers this means knowing which facts matter for a client, when they matter, and what to do about them.

Clients clearly feel the same way. In CFS research^[4] released in March 2026, 83 per cent of Australians surveyed said access to human interaction matters when it comes to their finances, particularly for super and long-term investment decisions. CFS interpret this as meaning client confidence comes from working with an adviser who understands their personal circumstances, applies judgement, and is accountable for their advice.

How AI is changing the advice conversation

Advisers know clients are already using AI alongside advice.

In 2026 the US-based Financial Planning Standards Board (FPSB) conducted a survey of 8,095 planners across 37 territories. Respondents said the most common usage by their clients was for general financial education (33 per cent), and a quarter said clients use it to validate or compare professional advice^[5].

According to that research, planners saw two main effects on their relationship with clients. The most common was a greater need to put AI-generated information in context or correct it (21 per cent). Nearly as many (19 per cent) said client use of AI had led to more collaborative discussions.

For some advisers, however, client usage of AI may also create relationship risks.

In four experiments^[6], Gerri Spassova of Monash University and Mauricio Palmeira of the University of South Florida examined how advisers reacted when clients had consulted AI. The first experiment involved 200 professional financial advisers, responding to a hypothetical client scenario; in the other three, participants imagined themselves as advisers. Participants reported less motivation to keep working with a client who had consulted AI than with one who had sought advice from another adviser. The effect remained even when the AI had been used only for background information. The researchers suggest advisers may interpret the comparison as insulting their professional identity.

Whatever the reaction, the question is the same. If the client already has an answer, what does the adviser add to it?

Judgement begins where the answer ends

As Russell notes, AI can research a question and model scenarios in seconds. What it cannot do, their 2026 report suggests, is help a client work through a terminal illness diagnosis with their family, or supply the judgement needed when life’s biggest decisions are also its most emotional.

Researchers at the University of Canberra, Bomikazi Zeka and Raechel Johns, put hypothetical financial scenarios, involving consumers experiencing vulnerability, to three AI tools: ChatGPT, Claude and Perplexity^[7]. The tools responded in different ways, but all were inconsistent in considering the vulnerabilities in the scenarios. Perplexity often referred users to professional advice but gave little detail. ChatGPT avoided implicit bias but did not adequately address the consumer’s vulnerabilities. Claude gave comprehensive recommendations but leaned towards self-guided planning. Each answer left something important for an adviser to provide, leading the authors to conclude that AI should serve to complement – not replace – professional financial advice.

That points to two questions an adviser must put to any AI answer. Does it fit this client’s goals and circumstances? And what – because nobody asked the tool – has it left out?

Some decisions have no single correct answer, because each alternative protects one thing the client cares about but puts another at risk. For example, holding cash gives quick access to funds if needed, but its purchasing power erodes over time; repaying debt lowers commitments (and the stress that can accompany them) but may leave less for other immediate plans.

Trade-offs are common in advice, and one of the specific areas where advisers add value. Indeed, this is reflected in Russell’s proprietary Value of an Adviser formula, as item ‘C’ (Choices and trade-offs).

When the adviser uses AI

According to Russell, around nine in 10 advisers now use AI in some form within their business, primarily to prepare client meeting notes and summaries (73 per cent), check or validate information (39 per cent), and generate documents and reports (37 per cent). More than 80 per cent say AI has had a positive impact on their business.

Of course, the questions an adviser puts to a client’s AI answer should apply equally to the adviser’s own use of it. FPSB’s 2026 survey suggests firms are still working out how to review AI outputs. Accuracy and reliability of outputs were among planners’ top two concerns (48 per cent), yet only 23 per cent of firms had a comprehensive AI policy and 22 per cent had no guidance at all. Among firms with a policy, 48 per cent said it covered human review before AI outputs reach clients^[8].

Canada’s CFP certification body, FP Canada, issued guidance for planners encouraging the disclosure of AI use in engagement letters or deliverables, specifying where AI contributed and noting that the planner has reviewed the final recommendations^[9].

Clients certainly expect the same openness. In a TAL and McCrindle survey of 2,005 Australians, 57 per cent were comfortable with their adviser using AI tools to inform advice^[10], and about 85 per cent wanted to know if advice from a human adviser had been generated by AI^[11].

For advisers, the practical question is less whether to use AI than what to check before relying on it, and whether the client can see that they did those checks. 

Case study: from AI answer to advice decision

Illustrative example.

Mara, 57, is an architect with a substantial super balance and a mortgage she has nearly paid off. Before meeting an adviser for the first time, she asks a popular AI tool whether she can afford to retire at 60, plugging in her balances, her spending needs and her retirement date. The answer is clear and confident: she can sustain her planned spending if she keeps working full-time for another three years. She takes it to the meeting with her new adviser, saying it sounds sensible, but it is not quite what she wants.

While the AI tool answered the question it was asked, using the information it was given, it could not know the missing context. Mara is tired of the demands of her role but not sure she wants to stop working altogether. She would like more time to travel and is open to consulting. She had not mentioned any of this, and she did not know that a Transition to Retirement (TTR) pension existed, let alone that it could let her draw on her super from 60 while still working. So, she did not ask about it, and the AI tool did not raise it. Mara had not held anything back. She simply did not know what was relevant.

The adviser does not dispute the answer or make light of Mara’s use of AI. The answer it gave is plausible, but it has answered a narrower question than the one Mara was really asking. By asking what stepping back would let her do that she cannot do now, and what she would not be prepared to trade off, the adviser defines the real decision: not just whether Mara can afford to stop working at 60, but how much of her working life she is willing to give up.

The adviser models an alternative: reducing her hours now, with some consulting, and starting a TTR pension as soon as she is able. During the three years until she turns 60, she will need to live off reduced earnings, and any savings outside super, so the projection is tested against weaker investment returns, and against shortfalls in consulting income.

Staying full-time gives the larger buffer. Reducing her hours gives Mara more time now and still supports her planned spending, but with a smaller reserve. The adviser sets out what each path assumes and what it gives up and asks which outcome Mara most wants to protect. She chooses to reduce her hours, with a review after a year. Weighing her competing priorities with her is the Guide role Russell describes.

The file notes record the options considered and the assumptions behind each.

Making judgement visible

In a world where information is easier to find than ever before, human judgement and expertise are the true value added by financial advisers.

Russell’s 2026 Value of an Adviser research reinforces this, finding expertise – the ability to apply judgement when the answers aren’t obvious – to be the strongest driver of client satisfaction. CFS’s research points the same way: 83 per cent of Australians surveyed said access to human interaction matters when it comes to their finances. At the same time, while consumers are using AI, this does not appear to be diminishing their appetite for advice. The proportion of non-advised investors who say AI makes them more likely to seek advice is higher than the proportion who say it makes them less likely.

Mara’s case study illustrated how professional adviser judgement is not about dismissing the AI or criticising her decision to use it, but about understanding the client’s broader context and investigating a more comprehensive list of options. The AI tool could not answer questions Mara had not asked, and she did not know what she did not know. Her adviser used judgement to get to the heart of the question she was really asking.

Advisers are using AI too, and most clients are comfortable with that. But about 85 per cent want to know if advice from a human adviser has been generated by AI, so advisers should question their own AI outputs as rigorously as they would a client’s and let the client see that this was done. In an AI world, judgement and expertise remain the human advantage, and advisers need to make their application visible.

Take the FAAA accredited quiz to earn 0.5 CPD hour:

CPD Quiz

The following CPD quiz is accredited by the FAAA at 0.5 hour.

Legislated CPD Area: Client Care & Practice (0.5 hrs)

ASIC Knowledge Requirements: Client Engagement (0.5 hrs)

please log in to start this quiz

Read the series:

CPD: A practical system for building client confidence – Value of Advice (Part 1)

———–

References:
[1] https://russellinvestments.com/au/support/financial-adviser/business-solutions/practice-management/value-of-an-adviser
[2] Ibid.
[3] Ibid.
[4] https://www.cfs.com.au/about-us/media/cfs-tech-AI
[5] https://fpsb.org/impact-of-ai-on-financial-planning-global-research/
[6] https://www.sciencedirect.com/science/article/pii/S0747563226000312
[7] https://doi.org/10.1002/cfp2.70040
[8] https://fpsb.org/impact-of-ai-on-financial-planning-global-research/
[9] https://fpcanada.ca/professional-responsibility/practice-guidance/responsible-use-of-generative-artificial-intelligence–guidance-for-the-profession
[10]https://www.insurancebusinessmag.com/au/news/technology/ai-enters-financial-decisions-as-australians-still-turn-to-human-advice-590107.aspx
[11]https://newshub.medianet.com.au/2026/09/more-information-more-tech-but-australians-still-lack-financial-confidence/171320/

The post CPD: The human advantage – making your expertise visible in an AI world – Value of Advice (Part 2) appeared first on AdviserVoice.