Where the financial promotion line actually sits, side-by-side breach versus compliant rewrites, the phrases that flip a post into a promotion, and a pre-publish checklist you run on every post before it goes live.
A LinkedIn post becomes a financial promotion under FSMA Section 21 when it invites or induces investment activity. General education and factual commentary that do not nudge the reader to invest usually stay outside the rules. The practical safeguards: a banned-phrase list, the education-versus-advice line, and a pre-publish check on every post.
The line is not the topic. It is whether the post invites or induces investment activity. Stay in general education and factual commentary and most posts sit outside the financial promotion definition. Cross into product nudges, performance claims, or outcome language and you trigger the fair-clear-not-misleading standard, risk warnings, and approval rules. This guide gives you the line, breach-versus-compliant rewrites, a banned-phrase list, and a checklist. Your compliance officer makes the final call.
Most advisers think FCA compliance on LinkedIn is about avoiding certain topics. It is not. You can post about pensions, ISAs, investing, tax, and retirement all day. The line is whether your post is a financial promotion.
Under FSMA Section 21, a communication is a financial promotion if it invites or induces a person to engage in investment activity. Read that twice. The trigger is the invitation or inducement, not the subject. A post that explains how an ISA works in general terms is unlikely to be a promotion. A post that says "open an ISA with us before April" is heading straight into one.
If a post is a financial promotion, it must be communicated or approved by an authorised person, be fair, clear and not misleading under COBS 4.2, and carry any required risk warnings. For most advisers, the simplest path is to write posts that never become financial promotions in the first place. Stay in education and commentary, and the heavy requirements do not apply.
The same idea shows up a second way. General education is information that helps anyone understand a concept. Advice is a recommendation tailored to a specific person's circumstances. LinkedIn is a public feed, so everything you post is general by definition. The risk is language that makes a general post feel like a personal recommendation.
"Here is how the pension annual allowance works" is education. "You should top up your pension before the tax year ends" reads as advice to whoever is reading. Same topic, different side of the line. Keep the verb general. Explain, describe, set out the trade-offs. Do not instruct the individual reader to act.
The fastest way to internalise the line is to see it. Each pair below takes a common adviser post and rewrites it from breach territory to compliant.
The first version has a performance claim, an outcome promise, and an investment-activity nudge. Three breaches in two sentences. The second explains a concept and invites understanding, not action.
The first uses a comparative superlative ("the best pension") and an outcome claim ("losing money"). The second names no product as best, explains the concept, and points back to individual circumstances.
The first presents a prediction as certainty and instructs a specific action. The second offers a considered, factual perspective with no instruction and no forecast dressed up as fact.
The first promises guaranteed returns, implies capital is risk-free, and promotes a specific product. The fourth is the kind of post that draws regulatory attention fast. The compliant version educates a reader on what to scrutinise, naming no product and promising nothing.
These phrases and patterns tend to flip a post into breach territory. Treat the list as a hard filter. If a draft contains any of them, rewrite before it goes near publish.
If you use ChatGPT or Claude to draft posts, the banned-phrase list becomes a system instruction, not a memory test. Raw AI output is the single biggest compliance risk, because generic models default to exactly the marketing register that breaches: outcome language, benefit promises, and calls to invest.
The fix is to build the list into the prompt so every draft starts inside the rules. A workable instruction block looks like this.
"You write LinkedIn posts for a UK financial adviser under FCA rules. Never use performance figures, outcome promises, comparative superlatives, or calls to invest. Never name a specific product or fund as a recommendation. Never instruct the reader to take an action with their money. Explain concepts generally. Note where things depend on individual circumstances. If a topic cannot be covered without breaching, say so instead of writing the post."
Pair that with a voice prompt so the output sounds like you, not a compliance robot. The method for the voice side is in how to make ChatGPT sound like you, and the structure in how to build a voice prompt. Used together, the AI removes the blank page while the instruction block keeps the draft inside the rules. The human check still runs last.
Run this on every post before it goes live. It takes under a minute once it is a habit, and it catches the breaches the banned-phrase list does not.
1. Does this post invite or induce investment activity? If yes, rewrite or route through approval. 2. Any performance claim or figure? Remove it. 3. Any outcome promise or guarantee? Remove it. 4. Any comparative superlative? Remove it. 5. Does it read as a recommendation to an individual? Make it general. 6. Where your firm requires it, has your compliance officer signed off? Then publish.
The checklist is the gate. AI speeds up everything before it, but nothing skips it. For the wider content system this slots into, the four post types and the calendar, see LinkedIn content for financial advisers UK.
The Voice System Playbook shows you how to set up an AI that drafts in your voice with the compliance instruction block built in, so every draft lands inside the rules before you even edit it. Free, no call required.
Get the Voice System PlaybookThis page handles the regulatory mechanics of a single post. For the full content engine, the post types, the calendar, the prompts and the weekly batch, read LinkedIn content for financial advisers UK. For the broader marketing approach across UK and US regimes, read marketing for financial advisers. And for the whole UK done-for-you landscape and where regulated professions fit, the pillar is done-for-you LinkedIn content in the UK.
Every regulated UK profession faces a version of this problem: a content opportunity boxed in by a regulator. The system transfers; only the rulebook changes.
Under FSMA Section 21, a communication is a financial promotion if it invites or induces a person to engage in investment activity. The trigger is invitation or inducement, not the topic. A post explaining how compound interest works in general terms is unlikely to be a financial promotion. A post that nudges the reader towards buying a specific product, fund, or service in a way that encourages investment activity can be. If a post is a financial promotion, it must be communicated or approved by an authorised person, be fair, clear and not misleading, and carry any required risk warnings. The safest LinkedIn content for most advisers stays in general education and commentary that does not invite investment activity.
It is high risk and usually best avoided in organic LinkedIn content. Naming a specific fund or product alongside language that encourages a reader to invest tends to turn a post into a financial promotion, which then triggers the fair, clear and not misleading standard, prominent risk warnings, and the approval requirements. Most advisers keep specific product discussion inside the advice relationship, where suitability and documentation apply, rather than on a public feed. General concept education without product names carries far lower risk and still builds authority.
No. A risk warning is required where a post is a financial promotion and the rules call for one, for example prominent warnings under COBS 4.6 or the specific requirements for high-risk investments under COBS 4.5A. A general education post that is not a financial promotion does not need a standard investment risk warning. The practical rule for advisers: if a post drifts towards encouraging investment activity, either rewrite it so it does not, or treat it as a financial promotion with the full requirements applied. When in doubt, your compliance officer makes the call.
Only if you publish raw AI output. Generic AI defaults to marketing language that breaches financial promotions rules: outcome claims, benefit promises, comparative superlatives, and investment-activity nudges. The compliant approach is to give the AI a banned-phrase list and clear instructions up front so the draft starts inside the rules, then apply a human compliance check before publishing. Used this way, AI reduces compliance risk by making the constraints explicit on every draft, rather than relying on memory. The model speeds up drafting; it never replaces the compliance gate.