The client has already asked the AI. Now the advice has to be better.

In summary

• Vanguard’s global chief economist has described ChatGPT as probably the largest provider of financial advice in the world.
• Clients increasingly arrive at an adviser having consulted AI first, with sharper and more specific questions.
• The basic information layer of advice is heading towards zero cost. Value moves to context, judgement and advice that stands up to scrutiny.
• Quality paraplanning is where a client’s real circumstances meet the strategy.

The client now arrives informed, and sometimes misinformed

Earlier this year, Vanguard’s global chief economist Joe Davis told ifa that ChatGPT is probably the “largest provider of financial advice in the world”, and almost certainly so in the United States. Nobody has firm numbers. ChatGPT itself will not offer one. Anyone working in advice, though, will recognise the shift he is describing.

Davis called it both a threat and an opportunity. Six months on, both halves of that description are playing out in adviser meeting rooms.

Independent adviser Nathan Fradley, writing in ifa, has noticed a rise in enquiries from people who consulted their preferred AI before making contact. The giveaway is the precision of the questions. Sometimes the AI has referred them on. More often the questions open into a much bigger conversation.

His observation about the quality of that AI guidance is the one that matters. The information is often correct in isolation and wrong for the person, because it lacks the broader consideration that makes advice appropriate. He also worries about the people who never get in touch, and the outcomes that will surface years from now.

The dynamic in the room has changed as a result. A client who once arrived with a vague sense that they should be doing something now arrives anchored to a specific answer. The adviser is no longer starting from a blank page. They are starting from a plan the client half trusts, and the first job is to work out how much of it holds.

Basic information is heading to zero, so value moves

Davis frames the economics plainly. Lower the barrier to information and history is clear: the price of an identical service goes to zero. A generative tool has removed much of the intimidation that once stopped people asking about money at all.

The question of trust remains. A recommendation from an AI is easy to obtain. Whether it holds up as something a client can rely on with the tax office, or as anything resembling a legal opinion, is a different matter. Davis expects the value of advice to survive, concentrated in the fundamental process behind it, and he expects clients to demand a higher level of service because the basics are now easy to get.

The bar on advice rises with the client

This is where threat and opportunity meet. The informational part of an adviser’s value proposition is being commoditised. The rest of it is becoming more valuable, provided the adviser can deliver it.

A client who has done their AI homework arrives engaged, with specific questions and often a rough plan. The adviser has to meet that with something more rigorous than the AI offered: a strategy tested against the client’s full circumstances, and a clear account of why it differs from what the machine suggested.

Davis makes a related point about where AI becomes transformative for a practice. Saving an hour on the plan is the minimum. The real gain lies in richer advice that draws in estate planning, charitable giving and more of the client’s life, and richer advice is more complex advice to build.

Why the work behind the advice gets harder

The gap Fradley describes, between information that is right in a vacuum and advice that is right for the person, is exactly the gap that advice construction closes. It closes through research, modelling, comparison of alternatives, and documentation that explains why the recommended path suits this client where the AI-generated one falls short.

Consider a common case. An AI tells a client in their late fifties to maximise salary sacrifice and open a transition to retirement pension. On its own, that is reasonable. It says nothing about the spouse’s super balance, the insurance held inside the fund, a Centrelink position two years away, or a parent’s aged care decision that will change the family’s cash flow. The recommendation was accurate. The advice was incomplete.

When a client walks in with a ChatGPT-shaped plan, the Statement of Advice has to answer it. Vague reasoning and template-heavy documents will not survive that conversation. The strategy has to be visibly built on the client’s own numbers, and the reasoning has to be laid out in a way an informed, sceptical client can follow.

That standard is rising at a time when licensees remain cautious about AI drafting advice documents. Most have yet to approve a tool for the task, and the adviser and licensee own the advice however it is produced. The construction stays human, specialist work.

Where paraplanning fits

Quality paraplanning is the broader consideration made concrete. It takes the adviser’s strategy and the client’s real circumstances and turns them into advice that is modelled, compared, compliant and clearly explained. As clients arrive better informed and more demanding, that discipline decides whether the adviser’s answer lands.

At Mutual Plans, that is our work. AI has made the basic answers cheap. It has made the good ones worth more.

Frequently asked questions

Is AI replacing financial advisers?

No. Generative AI is commoditising basic financial information, which raises the value of personal advice that accounts for a client’s full circumstances and can be relied on.

What should an adviser do when a client arrives with AI-generated advice?

Treat it as the start of a bigger conversation. Test it against the client’s complete situation, then document clearly why the recommended strategy differs and why it suits them.

Can AI produce a Statement of Advice?

Most licensees have yet to approve a tool for the task. The adviser and licensee remain accountable for the advice however it is produced, and human review and sign-off are still required.

Glossary

Generative AI

AI systems, such as ChatGPT, that produce text, images or other content in response to a prompt.

Personal advice

Financial advice that takes into account an individual’s objectives, financial situation and needs, and carries specific legal obligations.

General advice

Information or guidance that does not consider an individual’s personal circumstances.

Best interest duty

The legal obligation on advisers to act in the best interests of their clients when providing personal advice.

Statement of Advice (SoA)

The document that sets out the personal advice given to a client, including the strategy and the reasons behind it.

Paraplanning

The research, modelling, analysis and documentation that sits behind financial advice before it reaches the client.