The FCA's Balancing Act: Cracking Down on AI Advice While Opening the Door to Captive Insurance

The Financial Conduct Authority is doing two very different things at once. It's warning young investors that AI chatbots aren't the safety net they think they are, and it's helping design a brand-new insurance market from scratch. On the surface, these look unrelated. They're not.

Both moves reveal how the FCA sees its job right now: protect consumers from things that look official but aren't, while actively building infrastructure for things that don't exist yet. That's a harder balancing act than it sounds.

The AI trust problem nobody asked for

New FCA research on 18- to 40-year-old investors found something that should concern anyone who thinks regulation keeps pace with technology. Four in five less experienced investors have used AI to help with investing decisions. More strikingly, 56% of them trust AI tools more than TV and radio (47%), the press (46%), or social media influencers (29%).

That's not a fringe behaviour. That's the mainstream.

The real issue isn't that people use AI for research — that's fine. The issue is what they believe about it:

44% mistakenly think AI-generated financial information is regulated.
38% believe it's fine to make an investment decision based solely on AI output.
32% wrongly assume they'd be compensated by the FSCS or the Financial Ombudsman Service if AI advice went wrong.

None of that is true. General-purpose AI chatbots sit outside the FCA's regulatory perimeter entirely. If a chatbot gives you a bad steer and your investment tanks, there is no scheme, no ombudsman, and no compensation waiting for you.

To be fair, the same research shows people aren't completely naive. 73% know AI can produce inaccurate information, and 86% understand they should check sources. The gap isn't awareness — it's assuming protection that doesn't exist.

Lucy Castledine, the FCA's director of consumer investments, put it plainly: AI can help you research and understand jargon, but "you need to understand how you're protected and continue to use your own judgement." The regulator's InvestSmart website exists specifically to fill that gap.

Five rules the FCA wants investors to actually follow

Stay in the driving seat. AI can inform a decision. It shouldn't make it.
Check your sources. Ask the AI where the information came from, then verify independently.
Know there's no safety net. Unregulated AI advice means unregulated risk.
Past performance isn't a forecast. AI can only hand you historical data, not predictions.
Think long-term. A tip from a chatbot carries the same weight as one from a stranger at the pub — treat it accordingly.

This is the FCA doing what regulators do best: naming a risk clearly, without pretending the underlying technology is the enemy.

Meanwhile, a new insurance market is being built

At almost the same time, the FCA and the Prudential Regulation Authority jointly proposed something far more constructive: a new captive insurance regime designed to make the UK competitive in a fast-growing global market.

Captive insurance — where a company sets up its own insurer to cover its own risks — has traditionally been done offshore, in places like Bermuda or Guernsey. The UK has largely watched from the sidelines. This proposal is an attempt to change that, positioning Britain as a serious base for a market segment it currently has almost no share of.

This isn't a defensive regulatory move. It's an offensive one. The FCA and PRA are effectively saying: we want this business here, and we're willing to build the rules to attract it.

Two moves, one underlying logic

At first glance, warning consumers about AI and courting captive insurers look like they belong in different departments. They don't. Both reflect the same regulatory instinct.

Where unregulated technology is quietly replacing regulated advice, the FCA is stepping in to correct the record before real damage is done.
Where regulated infrastructure is missing and costing the UK economic opportunity, the FCA is stepping in to create it.

In both cases, the regulator isn't reacting to a crisis. It's trying to get ahead of a trend — one demographic (young, AI-comfortable investors), one structural (a market the UK has been losing to other jurisdictions for years).

That's arguably a more coherent strategy than it first appears. The FCA isn't anti-innovation. It's anti-unmanaged innovation. AI chatbots giving informal financial advice with zero accountability is unmanaged risk. A captive insurance market with no UK regulatory home was unmanaged opportunity. Fix both, and you get a financial system that's simultaneously safer for consumers and more competitive for business.

What this means if you're an investor or a business

If you're a younger investor using AI to research stocks or funds, the message is straightforward: use it, but don't outsource judgement to it. Nothing about a chatbot's confidence level tells you whether its output is accurate, current, or appropriate for your circumstances.

If you're a business considering captive insurance, the message is equally direct: the UK is trying to make itself relevant in a space it previously ignored. Whether that translates into real uptake depends on how the final rules land — but the intent is clear.

Key takeaways

AI is not regulated advice. Nearly half of young investors wrongly believe it is, and there's no compensation scheme if it goes wrong.
Trust in AI is already mainstream among 18–40 year olds, outpacing trust in TV, press, and influencers.
The FCA's five rules — stay in control, check sources, know there's no safety net, don't mistake history for prediction, think long-term — are the practical bare minimum for using AI safely.
The UK is actively building a captive insurance market, aiming to compete with established offshore hubs.
Both initiatives share one logic: correct unmanaged consumer risk, and create regulated space for unmanaged economic opportunity.