Skip to content

AI Search Is Eating Your Clicks. FCA Rules Still Apply

Sotiris SpyrouUpdated on

Share this article

LinkedInXEmail
AI Search Is Eating Your Clicks. FCA Rules Still Apply

If your buyers research mortgages, pensions or business banking inside ChatGPT or a Google AI Overview, the click you used to win is disappearing. Pew found people click a traditional result on just 8% of searches that show an AI summary, against 15% when none appears, and only 1% click a link inside the summary itself. For a regulated financial brand, the new game is being the source the AI quotes, done inside FCA financial promotion rules. That's harder than ranking, and most of your competitors haven't started.

This is the shift financial services CMOs keep underrating. The visibility problem and the compliance problem are the same problem now. Get cited by an AI engine without controlling how it describes your product, and you've published a financial promotion you didn't write. Here's what changed by June 2026 and what to do about it.

What's actually changing in how buyers find financial products?

People used to type "best ISA rates" and scan ten blue links. Now they ask a full question and read one synthesised answer. That answer is built from sources the engine trusts, and it often ends the search before anyone reaches your site.

The traffic effect is measured, not theoretical. The Pew Research Center study from July 2025 tracked the browsing of 900 US adults. Their findings:

What Pew measured With an AI summary Without one
Clicked a traditional search result 8% of searches 15% of searches
Clicked a link inside the AI summary 1% of all visits n/a
Ended browsing after the page 26% 16%

Around 18% of searches in the study produced an AI summary. That share is rising, and financial queries are exactly the kind of high-consideration questions these systems try to answer in full.

For a bank or insurer, the read is blunt. The page that used to earn the click now has to earn the citation. If the AI answer names three providers and you're not one of them, you're invisible at the moment of research, and you'll never see it in your click data.

Why can't financial services CMOs just do normal SEO for this?

Because the work that gets you cited is different from the work that gets you ranked, and because everything you publish is a financial promotion.

On the visibility side, the peer-reviewed evidence is clear that AI engines reward different things than classic search. The GEO: Generative Engine Optimization paper, accepted to KDD 2024, found that content tuned for generative engines can lift its visibility in AI responses by up to 40%. The tactics that worked were adding cited statistics, quotations from credible sources, and clear authoritative language. Keyword stuffing, the old SEO reflex, did almost nothing. We unpack this in our GEO vs SEO guide.

That last point matters more for finance than any other sector. The signals an AI engine uses to decide who to trust, citations, named sources, transparent claims, accuracy, are the same signals a regulator wants to see in a financial promotion. Done right, your compliance discipline becomes your visibility edge. Done badly, you're optimising for citations in ways that breach the rules.

What FCA rules govern AI-era financial marketing?

The core rule predates AI and still binds every word an AI engine might quote from you. Under COBS 4.2, every financial promotion must be fair, clear and not misleading. The FCA is explicit that a promotion built entirely from accurate statements can still be misleading if the overall impression, or what it leaves out, points the customer to a false conclusion.

Two things make this sharper in the AI era.

First, you lose control of framing. When an AI engine pulls a sentence from your page and drops it into an answer next to a rate, a risk warning that lived two paragraphs down may not travel with it. If the engine quotes your headline benefit and strips the caveat, the impression the customer receives can breach the fair, clear and not misleading standard, even though your original page was compliant. That's a real exposure, and it's new.

Second, Consumer Duty raised the bar. The Duty, in force since July 2023, requires firms to actively support customer understanding, not just avoid misleading people. An AI summary that confuses a buyer about your product is a Consumer Duty problem, whoever generated the words.

On AI specifically, the FCA has not written separate AI rules. Its stated approach is "same activity, same risk, same rule," set out in the AI Live Testing feedback statement FS25/5. Translation for marketers: there's no new rulebook to hide behind and no special dispensation. The promotion rules you already follow apply to AI-mediated discovery in full.

How should a regulated brand get cited without breaking the rules?

Four moves, in order. None of them require a new platform. They require editorial and compliance discipline applied to content.

  1. Write self-contained, quotable statements. Assume any sentence can be lifted on its own. Pair every benefit with its risk warning in the same breath, so the caveat travels when the claim does. This is good GEO and good compliance at once.

  2. Cite your sources visibly. The GEO research shows named statistics and credible citations lift AI visibility. The FCA wants substantiated claims. Put the source next to the number. One action, two wins.

  3. Answer the whole question. Buyers now ask "should I overpay my mortgage or invest the difference," not "mortgage rates." Pages that genuinely answer the complete question get cited. Thin keyword pages don't.

  4. Monitor what the engines say about you. Check monthly what ChatGPT, Google AI Overviews and Perplexity actually tell buyers about your products. Treat a misleading AI description of your product as a financial promotion issue, because under the fair, clear and not misleading standard, it is. Most firms have no process for this. That gap is your opening.

This is the responsible version of AI visibility work. No dark patterns, no manufactured authority, no claims you can't substantiate, because in regulated finance those tactics aren't just risky, they're a breach waiting to be found. Our complete guide to AI engine optimisation sets out the wider method.

Frequently asked questions

Do AI Overviews really cut clicks to financial websites?

Yes. The Pew Research study from July 2025 found users clicked a traditional result on 8% of searches showing an AI summary, against 15% without one, and clicked a link inside the summary on just 1% of visits. Financial queries are high-consideration questions these engines try to answer in full, so the effect is likely to be felt strongly in this sector.

Does the FCA have specific rules for AI in marketing?

No separate AI rulebook. The FCA applies "same activity, same risk, same rule," set out in FS25/5. Your existing financial promotion duties under COBS 4.2, fair, clear and not misleading, plus Consumer Duty, apply fully to AI-mediated discovery.

Can an AI engine misquoting my page count as a non-compliant promotion?

It's a real risk worth managing. If an engine lifts a benefit from your page and drops the risk warning, the impression a customer receives can fail the fair, clear and not misleading test even though your page was compliant. Get your own legal and compliance view, and build content so caveats travel with claims.

What actually makes content show up in AI answers?

The GEO paper found cited statistics, quotations from credible sources, and clear authoritative writing lifted visibility in AI responses by up to 40%. Keyword density barely moved the needle. For finance, the trust signals AI engines reward overlap closely with what a regulator wants to see.

The bottom line

The financial brands that win the next two years won't be the ones with the biggest ad budgets. They'll be the ones whose content is accurate enough to cite, clear enough to quote safely, and built so the risk warning never gets left behind. That's my firm view: in regulated finance, compliance discipline is the visibility moat, not the cost of doing business. Your competitors are still treating the two as separate departments. While they argue about whose budget this is, the AI answer to your buyer's question is being written without you in it.

Start by checking what ChatGPT and Google's AI Overviews tell a buyer about your flagship product today. If the answer is wrong, or missing, you already have your first project.

For the regulated-industry view of the wider risk picture, see our analysis of AI threats in banking and financial services.

Part of VerityAI's AI search compliance for financial marketers.

This is the kind of work our AEO services handles.

Share this article

LinkedInXEmail
Sotiris Spyrou - Author

Sotiris Spyrou

Sotiris Spyrou is the founder of VerityAI, a Responsible AI advisory for boards and AI-deploying businesses. With 27 years across agencies, global in-house roles, and the C-suite, he advises leaders on AI governance and risk, and on answer-engine visibility engineered without the dark patterns the rest of the industry is getting penalised for. He is the author of TRANSFORM, AI Moats, and Ethical AI.

Founder at VerityAI