Intelliflo’s 2026 survey found that 74% of financial advice firms now use AI in some part of their operations, up from 43% the year before. That is a fast move, but most of it sits at Level 1: a prompt here, a template there, a ChatGPT subscription a couple of advisers use on their own initiative. An AI-enabled firm is a different thing. AI is built into its workflows, overseen properly, and gives the team measurable time back.

The gap between the two can be closed in twelve months with a plan. This is the plan.

Why a phased approach works better than a pilot

Most firms that stall on AI try to do too much at once. They buy a tool, run a pilot, get inconsistent results, and conclude that AI “isn’t ready” for their firm. What they hit was an implementation problem, and the tool took the blame.

The second reason to phase it is the regulator. The FCA has been clear that AI in regulated firms must be governed, auditable, and human-supervised[1]. Firms that put AI on top of broken or undocumented processes end up with broken, undocumented AI. The sequence below builds the foundation first.

The question to keep asking is “where is human time spent on work that AI can support, with a person still responsible for the outcome?”, rather than “what can AI do?”

Months 1 and 2: map your workflows before you automate any of them

The first two months are about your processes, not AI.

  • Run an honest workflow audit. Walk through a client from initial enquiry to annual review. Write down every step. Note who does it, how long it takes, and whether there is a documented process or whether it lives in someone’s head. Pay particular attention to the handoffs: between adviser and paraplanner, between paraplanner and compliance, between back office and client-facing teams.
  • Find the highest-friction points. These are usually meeting preparation, suitability report drafting, client communication, compliance checking, and data entry across systems that don’t talk to each other. You are looking for work that is high-volume, fairly structured, and eating time that could go to clients.
  • Check your data hygiene. AI is only as good as what you feed it. If client records are filled in inconsistently, if the CRM is a mess, or if key documents live in a shared drive with no naming convention, fix that first. It is unglamorous, and it decides whether month 6 works.

By the end of month 2 you should have a written map of your top five workflow bottlenecks and a clear picture of where your data is and how clean it is. Everything after this builds on that document.

Months 3 and 4: build Level 1 prompting skill

Months 3 and 4 bring AI in, carefully, on two or three of the bottlenecks you mapped. Use tools your team already has, or cheap additions. Level 1 means education and prompting, not integration or custom builds. The aim is to build confidence and agree what good AI-assisted work looks like at your firm before you automate any of it.

The two most productive Level 1 wins for most advice firms are:

Meeting preparation and follow-up. A well-written prompt, working from your client record and the agenda, can produce a structured meeting brief in minutes. A prompt working from a transcript or notes can produce a draft follow-up letter or action list. Both save time, and neither takes adviser judgement out of the process.

First-draft suitability sections. AI can draft component sections of a suitability report from structured client data, and the adviser or paraplanner reviews, edits and is accountable for the final document. It does not draft the advice. It drafts the description of the client’s circumstances and the explanation of the recommendation, which the professional then checks against the actual advice position. The suitability-drafting playbook goes into this in detail.

One tool choice matters here. For regulated, client-facing outputs, use a grounded AI tool where you can. NotebookLM, for example, restricts its answers to sources you upload, which cuts the risk of the model adding information that isn’t in your files. Tools that can draw on their general training alongside your files, including ChatGPT Projects and Claude Projects in their standard set-up, are more likely to produce content that goes beyond your source material, and that distinction matters for compliance. Whichever tool you choose, a person reviewing every output before it reaches a client is still the main control. Pick the tool on purpose, not by default.

By the end of month 4 you should have two or three prompt workflows your team uses consistently, a clear policy on which outputs need human review before they go near a client (all of them), and a short internal log of where the tools have helped and where they haven’t.

Months 5 and 6: Level 2 integration where the evidence supports it

Months 5 and 6 connect the workflows that worked to your existing systems, so nobody is copying and pasting between them.

Level 2 integration usually means two or three tools talking to each other: your CRM, your document management system, and an AI step in between. It is configuration work, not engineering, and tools like n8n, Make or Zapier handle most of it. The builds typically take days to weeks and cost a few hundred to a few thousand pounds.

A realistic example for a mid-sized advice firm: a workflow that pulls a client record from the CRM, runs it through a structured prompt, and drafts a meeting brief into your document system for the adviser to review. Nothing goes to the client and nothing is finalised. A person is still at every step with consequences, but the work that used to take thirty minutes takes five.

If client onboarding is one of your bottlenecks, this is the phase to deal with it. Digital onboarding and KYC collection are integration problems of exactly this shape.

Months 7 and 8: put governance in place before you scale

Governance is the phase most firms skip, and skipping it is why many of them stall or run into trouble later.

By month 7 you probably have several AI-assisted workflows running. Before adding more, answer these questions formally:

  • Who is accountable for each AI-assisted output? In a regulated firm, “the AI did it” is not an answer the FCA will accept. Every suitability recommendation, client communication and compliance document has a responsible person. Your AI governance framework needs to name those people and make their accountability explicit.
  • How are you logging AI use? If the FCA asks what role AI played in a specific client outcome, you need to be able to answer. That means keeping records of which workflows used AI, which outputs were AI-assisted, and what human review took place. It doesn’t need to be elaborate. It needs to be consistent and easy to retrieve.
  • What is your review cadence? Providers update models, change terms, and sometimes change their products entirely. A workflow that works correctly in month 6 may behave differently in month 10 if the underlying model has changed. Make a quarterly review of your AI workflows a standing item.

A practical format: a one-page AI register naming each active workflow, the tool it uses, the person accountable for its outputs, and the date it was last reviewed. That page is your governance record. The governance pack checklist covers what else belongs alongside it.

Months 9 and 10: measure what has changed

By month 9 you should have enough data to be honest about what AI has and hasn’t delivered.

The metrics that matter for an advice firm are concrete: time per suitability report, from instruction to signed-off document; time per meeting cycle; new clients processed per paraplanner per month; and the error rate in client documents, measured by the number of compliance corrections raised. If the month 1 and 2 audit captured these figures, you now have a baseline to compare against. If it didn’t, start recording them now, because a vendor’s benchmark tells you about their test set and not about your firm.

This is also the phase to be honest about what hasn’t worked. If a workflow isn’t being used, find out why. Usually the time saving isn’t worth the change in habit, the output isn’t reliable enough, or the person accountable for the outcome doesn’t trust it. Each has a different fix, and none of them is “add more AI”.

Months 11 and 12: embed it and decide what comes next

The final phase is about embedding what you have built and deciding on purpose what Level 3 looks like for your firm, if anything.

Level 3, in my framework, means real engineering: custom pipelines, several agents working together, and complex logic that mixes fixed rules with model judgement. For most advice firms under 50 staff, nothing they need requires Level 3. The risk at month 12 is that a firm, having built confidence, reaches for complexity it doesn’t need.

The question at this stage is where the highest-value human time in the firm is still going on work AI could support. Twelve months of data will answer it, and that answer is where to focus next.

Two developments are worth tracking as you plan beyond month 12. Intelliflo has signalled that its next platform generation will be an agentic open platform that runs processes across firm systems, third-party tools and its own applications, so if your practice management runs on Intelliflo, your integrations are going to change. The FCA has also been consulting on reform of the Advice Guidance Boundary, with the aim of making room for different kinds of advice. Both are worth watching, not acting on yet.

Where to start

Start with the month 1 workflow audit. Not a technology decision, not a tool evaluation: a map of where your people’s time goes and where the friction is. Everything else follows from that, and the firms that get there in twelve months are the ones that map first, build on evidence, govern before they scale and measure honestly.

This article is for informational purposes only and does not constitute regulated financial advice or a compliance opinion. Consult a qualified compliance professional for advice specific to your firm.

Sources

[1] FCA, ‘Rethinking regulation in the age of AI’, Financial Conduct Authority, 26 June 2026. Available at: https://www.fca.org.uk/news/speeches/rethinking-regulation-age-ai