Financial advisers – your jobs have gone.You and your clients just don’t know it yet.

Financial advisers – your jobs have gone.
You and your clients just don’t know it yet.

That’s the uncomfortable starting point for this piece.

Mind the gap? What gap?

For years, the UK financial services industry and its regulators have talked about the advice gap – the millions who “need advice” but don’t get it.

The FCA’s latest Financial Lives work shows that less than 9% of adults received regulated advice on investments, pensions or retirement planning in the past 12 months.

At the same time, Lloyds Banking Group’s 2025 Consumer Digital Index reveals something far more interesting:

• 56% of UK adults – around 28.8 million people – have used AI in the past 12 months to help manage their money.
• Personal finance is now the number one use of AI in the UK.
• Among these users, six in ten say they use ChatGPT, making it the leading AI platform for money help.
• Around a third use AI for investment research and recommendations, a quarter for debt strategies, and almost four in ten for future planning such as pension questions.

Put those together and the headline writes itself:

The advice gap hasn’t been closed by regulators or product providers.
It’s being bypassed by the public – with LLMs.

When over half the population are already using AI as a de-facto “financial adviser”, and fewer than one in ten are using regulated advice, the old narrative collapses. The gap has already been jumped – just not in the way the industry expected.

---

What people are actually doing with AI

Look at the use-cases Lloyds highlights:

• Budgeting and day-to-day money management
• Savings planning and general financial education
• Investment research and ideas
• Debt management strategies
• Pensions and future financial planning

In parallel, press coverage shows people using ChatGPT (and similar tools) to:

• Compare tax options and business structures
• Decode HMRC rules and correct their own filings
• Draft complaints and challenge decisions
• Build their own plans when the quote from an accountant or adviser is unaffordable.

In other words, the public is doing – unprompted – exactly what the industry said they couldn’t do without us.

---

What LLMs can already do (today, for free)

Let’s be concrete about capability. Off-the-shelf LLMs – used intelligently – can already:

1. Decode complexity on demand

LLMs like ChatGPT, Claude and Gemini can:

• Translate pension and investment documents into plain English
• Summarise a 40-page policy into a one-page brief
• Compare options side by side (“Option A vs Option B – pros, cons, risks, likely long-term implications”)
• Explain tax rules and allowances at a level most humans never will

This demolishes one of the industry’s favourite myths: that the regulations are simply “too complex” for ordinary people.

2. Build lifetime cashflow forecasts in minutes

Give an LLM:

• Ages, earnings, contributions, state pension estimates
• Basic growth and inflation assumptions
• Planned life events (children, business sale, downsizing, care needs)

…and it can produce:

• Year-by-year projected income and expenditure
• Asset values and drawdown trajectories
• “What if?” comparisons (retire at 60 vs 67, work part-time, increase pension contributions, etc.)

Claude, in particular, is already very good at ingesting a rough fact-find and spitting out the equivalent of a lifetime cashflow model as a table – the sort of thing planners currently charge four figures for, often produced in licensed software that the client never sees.

3. Run Monte Carlo-style scenarios

These models can:

• Generate randomised return paths around a given expected return and volatility
• Run multiple “trials” and show the proportion in which the plan “succeeds”
• Stress-test retirement plans against shocks (market falls, inflation spikes, longevity)

Gemini, when paired with a spreadsheet or a simple prompt, can produce Monte Carlo-like simulations at a level of sophistication that – let’s be honest – exceeds what most retail investors ever see from a human adviser.

4. Act as a 24/7 planning co-pilot

LLMs don’t get bored or rush the meeting:

• They will answer a basic explanation request ten different ways until the user “gets it”.
• They never roll their eyes at a “stupid question”.
• They remember the conversation and build on it over time.

For many people, that’s a more psychologically safe environment than sitting opposite a well-dressed stranger in a glass office.

---

So what exactly is left of the traditional adviser value proposition?

The product-distribution model relied on three pillars:

1. Information asymmetry – “we know things you don’t.”
2. Access asymmetry – “we can get you products you can’t reach.”
3. Confidence asymmetry – “we can give you certainty you don’t have.”

All three are eroding:

• Information is now abundant, contextual, and conversational.
• Products are a tap away on low-cost platforms.
• Confidence can be scaffolded by tools that walk people through complex topics step-by-step, not just by reassuring tones in a meeting room.

This is why I say:

Financial advisers – your old job has gone.
You just haven’t updated the job description yet.

If your offer is essentially:

• “I’ll input your details into a proprietary cashflow package.”
• “I’ll pick you a model portfolio and rebalance it.”
• “I’ll explain the tax rules you could have read elsewhere.”

…then, in economic terms, you are competing with a near-zero-marginal-cost technology that is already in your clients’ pockets.

---

“But AI is often wrong!” – yes, and that’s the real point

Investigations by consumer groups like Which? show that general-purpose AI tools can give:

• Incorrect, misleading, or risky financial answers
• Outdated links or made-up references
• Over-confident recommendations that ignore personal nuance

Lloyds’ own research found that around 80% of users worry about misinformation and 83% about data privacy when using AI for money decisions.

For me, this doesn’t prove AI can’t be used for advice. It proves something else:

We don’t just have an “advice gap”.
We have a structural trustworthiness gap.

The public is already using AI. They know it might be wrong. They still use it because the alternatives feel inaccessible, expensive, conflicted, or all three.

So the real question isn’t:
“Will AI replace advisers?”

The real question is:
“Who will combine AI with a structure the public can trust?”

---

From product seller to structural steward

If you’re a financial adviser reading this, there are only two honest paths forward.

Path 1: Compete against AI

You can:

• Denounce AI as dangerous and unreliable
• Lobby for regulation that locks it out of retail use
• Try to defend your spreadsheet and your 1% AUM fee

This might buy you a few years. It will not win the long game. History is unkind to those who try to hold back a technology that the public has already adopted.

Path 2: Compete with AI – on structure and conscience

Or you can accept that:

• LLMs will soon do 80–90% of the technical heavy lifting – modelling, projections, tax comparisons, scenario planning.
• Your unique value is not in keying numbers into a black-box tool…but in:
  • Designing a structurally trustworthy planning environment
  • Holding space for client emotion, fear, grief, ambition
  • Helping people articulate meaning, priorities, and trade-offs
  • Protecting them from predators and structural exploitation
  • Teaching them how to use AI safely, critically, and powerfully

In that world:

• ChatGPT, Claude, Gemini & co. are your junior paraplanners, not your enemies.
• The client owns the plan, the data and the tools.
• You are no longer an intermediary standing between the client and the market – you are a coach, mentor, and structural steward standing with the client as they navigate it.

---

“What gap?” – reframing the whole debate

If:

• Less than 1 in 10 adults are accessing regulated advice each year,
• More than half the country is already using AI for money management,

…then the regulatory conversation about “closing the advice gap” by watering down advice standards to allow more product-led “guidance” is yesterday’s battle.

The public has already bypassed the bottleneck.

The new reality:

• There is no advice gap. There is an alignment gap.
• The question is no longer “How do we sell more advice?”
• It is “How do we make sure the tools people are already using are aligned with their interests, not with the interests of those extracting value from them?”

That is a structural design problem, not a sales problem.

---

So, what should advisers do on Monday morning?

Three practical steps.

1. Stop pretending clients aren’t using AI

Assume every client already has:

• Googled their question
• Asked an AI chatbot for a view
• Seen a TikTok or YouTube video with “advice”

Bring it into the open:

• “What have you already read or asked online?”
• “Do you use ChatGPT or other AI tools for money decisions? How has that been?”

If you can’t talk about AI comfortably, you’ve already lost authority.

2. Pivot from answers to architecture

Stop selling “I know the answer.” Start offering:

• “Let’s build a GAME Plan – a clear cycle of Goals → Actions → Means → Execution – and use AI to test it from every angle.”
• “I’ll show you how to use these tools safely, challenge their output, and integrate the best of it into your plan.”

You become the designer of the planning system, not the sole source of wisdom.

3. Price for human value, not product access

If AI can do the cashflow projections, portfolio modelling, and tax comparisons:

• Why is your fee still pegged to assets under management, a tiny sliver of the client’s total wealth?
• Why not price against the depth of transformation – the quality of decisions, the progress toward life goals, the reduction in anxiety?

Done-by-you (DIY with templates), done-with-you (coaching), and done-for-you (curated support) models can all sit on top of the same AI-enabled planning architecture.

---

Conclusion: adapt, or be quietly automated

Lloyds has just told the country, in black and white:

• The majority are already using AI for money.
• ChatGPT is the dominant platform.
• They’re using it not just for pocket-money budgeting, but for investment, debt, pensions and long-term planning.

The industry can respond in one of two ways:

1. Deny and defend, insisting that “proper advice” can only live inside a regulated sales architecture that the majority will never access.
2. Embrace and elevate, using AI to dismantle information asymmetry, reduce costs to near zero, and focus human expertise where it truly belongs: on structural trust, conscience, and human flourishing.

At the Academy of Life Planning, we’re unapologetically choosing the second path.

Mind the gap?
What gap.

The void has already been filled.
Now the real work begins: making sure it’s filled with empowerment, not extraction.

---

Join the Movement

If you believe financial planning should serve life — not the other way around — join us at the Academy of Life Planning.
Together, we’re replacing extraction with empowerment.

http://www.aolp.co.uk