Most small advice firms are not asking whether AI works. They are asking whether using it will cause trouble at the next FCA review or a file check by their network. The short answer: the regulator has said it will not write a separate AI rulebook, which puts the weight on the rules you already know. This guide covers what that means day to day for a directly authorised IFA, a small network member or a paraplanning team.
What does the FCA actually say about AI?
On its AI approach page, the FCA states plainly: "We do not plan to introduce extra regulations for AI." It describes its approach as principles-based and focused on outcomes. Check the page before you finalise your own policy, as regulators update these positions over time.
For an adviser, that cuts both ways. There is no AI licence to apply for and no list of banned tools. There is also no safe harbour. If an AI drafted letter misleads a client, it is treated the same as a misleading letter typed by hand.
Which existing rules matter most when advisers use AI?
The FCA's AI page points to Consumer Duty, which requires products to fit the target customer, offer fair value, communicate in a way that meets customers' information needs, and provide proper support. It also points to the accountability rules for senior managers as relevant to the safe use of AI.
In a small firm, that usually means one director owns AI use. They decide which tools are approved, what they are used for, and how output is checked. Write that down in a page or two. If a supervisor asks, a short, dated policy beats a vague answer.
Data protection sits outside the FCA but bites just as hard. The AI data protection checklist for UK businesses covers the UK GDPR side in more detail.
How widely is AI already used in UK financial services?
The Bank of England and FCA survey is worth reading for more than the adoption figure. The number to notice is that 46% of firms said they had only a partial understanding of the AI they use, against 34% who said they understood it completely. About a third of use cases were third-party tools.
That is the trap for small advice firms. Buying AI built into a back office platform is easy. Being able to explain what it does with client data, and how you check it, is the harder part.
Where does AI save advisers the most time?
Typical starting points for a two to five adviser practice:
- Meeting notes. Record with consent, transcribe, then have AI produce a structured summary mapped to your fact find headings.
- Suitability report drafts. Feed in the fact find, risk profile and adviser notes, and get a first draft in your house style.
- Annual review preparation. Summarise last year's file, changes in circumstances and outstanding actions before the meeting.
- Readability checks. Ask the tool to flag jargon and long sentences, which supports the Consumer Duty consumer understanding outcome.
- Internal file checks. Compare a file against your own compliance checklist and list gaps for a human to confirm.
Accountants and solicitors face a very similar pattern of regulated paperwork, and how AI is being used in accountancy and legal practices is a useful comparison.
What should an adviser never hand over to AI?
AI tools can sound confident while being wrong about tax allowances, product features or a client's own figures. A draft that says the client wants capital growth when the notes say income is worse than no draft at all, because it looks finished.
Vulnerable clients need particular care. AI can help spot signals in notes, such as a recent bereavement or a health diagnosis, but deciding how to adapt the service is a human call. Complaints follow the same logic, and AI in complaint handling explains where the line usually falls.
How should a small advice firm handle client data with AI?
A practical order of checks before approving any tool:
- Is there a business or enterprise plan, and does it exclude your data from model training?
- Where is data processed and stored, and is that covered by your privacy notice?
- Is there a signed data processing agreement?
- Can you switch off chat history or set retention limits?
- Does your PI insurer or network have a view on the tool?
Update your privacy notice if AI transcription is new, and ask for consent before recording meetings. Clients rarely object when told plainly why it helps.
Is more FCA guidance on AI on the way?
The Treasury Committee's report on AI in financial services argued that the lack of AI-specific rules left consumers exposed, and asked the FCA for comprehensive, practical guidance by the end of 2026. The FCA's stated view remains that its existing rulebook is flexible enough to cover AI.
For a small firm, the sensible move is not to wait. A short AI policy, a list of approved tools and a review log will slot into whatever guidance follows.
Should advisers buy an AI tool or build their own workflow?
Specialist adviser platforms bundle transcription, report drafting and compliance logging, which suits firms that want one supplier to answer to. General assistants are usually cheaper and more flexible but need your own prompts, templates and checking process. Build versus buy for small businesses sets out the trade-offs in more depth.
Whichever route you pick, run a two week pilot on a handful of live files, measure time saved per file, and keep a note of every error the checker caught. That log is your evidence for the policy, and for the regulator if asked.
What does a sensible first month look like?
Week one: choose the task, usually meeting notes, and approve the tool after the data checks above. Week two: pilot with one adviser. Week three: review every output against the original notes and log discrepancies. Week four: decide, update the policy, and brief the rest of the team. Keep the log. It is cheap insurance.
AI for Financial Advisers UK — FAQ
Does the FCA allow financial advisers to use AI?
Yes. The FCA says on its AI approach page that it does not plan to introduce extra regulations for AI, and that its approach is principles-based and focused on outcomes. That means AI is allowed, but the rules you already follow still apply in full. Consumer Duty still requires good outcomes for retail customers, and SM&CR still puts accountability on named senior managers. In practice, a small advice firm can use AI for meeting notes, file reviews and first drafts of letters, as long as a qualified adviser checks the output, the client data is handled lawfully, and the firm can explain how the tool is used and who signs it off.
Can AI write a suitability report for a client?
AI can produce a first draft from your fact find and meeting notes, which is where much of the time saving sits. It cannot take responsibility for the recommendation. The adviser still has to confirm the facts are right, the recommendation fits the client's objectives and attitude to risk, and the reasons are explained in plain language the client can follow. Treat the draft like work from a new paraplanner: useful, quick, and always checked line by line before it leaves the firm. Keep a record that a named person reviewed and approved the final version.
How many UK financial firms already use AI?
The Bank of England and FCA survey published in November 2024 found that 75% of responding firms were already using AI, with a further 10% planning to within three years. Those respondents were mostly larger regulated firms such as banks and insurers, so the figure does not describe small advice practices directly. It does show the regulator expects AI to be normal across financial services, and that questions about how you govern it are likely to come up during supervision or file reviews.
Is new FCA guidance on AI coming?
Possibly. In January 2026 the Treasury Committee published a report calling on the FCA to give comprehensive, practical guidance on how its consumer protection and individual accountability rules apply to AI, by the end of 2026. That was a recommendation, not a rule. Until any guidance lands, the safest position is to assume existing rules already cover your AI use, document how you use it, and review your approach when the FCA publishes anything new.
What client data should never go into a free AI chatbot?
Anything that identifies a client or their finances: names, dates of birth, National Insurance numbers, policy numbers, account balances, health details from protection cases, and full fact finds. Free consumer tools may keep prompts or use them to improve models depending on settings, which is hard to square with UK GDPR and client confidentiality. Use a business plan with data retention controls and a signed data processing agreement, or strip identifying details before pasting anything in.
