Copy article

The FCA Built a System That Punishes Advisors. Now the Advisors are Jumping Ship Leaving Unregulated AI and Scammers to Fill the Void

ended 26. March 2026

The FCA is spending £30,000 a month on Palantir to hunt for financial criminals. Meanwhile, the number of financial advisory firms has dropped 15.6% since 2022, under 25s entering the profession have fallen below 200, and more than half of UK adults are already using AI chatbots for financial advice with no regulatory oversight whatsoever.

The Palantir contract, defended before the Treasury Committee yesterday by FCA chief executive Nikhil Rathi, is supposed to help crack down on financial crime. Fair enough. But it arrives at a peculiar moment. The FCA's own Mills Review, launched in January, is openly asking whether AI systems are already providing services functionally equivalent to regulated advice while sitting entirely outside the regulatory perimeter. The regulator knows this is happening. It just hasn't done anything about it yet.

Here's the uncomfortable question: what happens when you regulate human advisers so heavily that they leave the profession, while the AI tools replacing them answer to nobody? Begbies Traynor data shows over 1,000 advisory firms have exited since 2020. Nearly a third of remaining advisers expect to retire within five years. The pipeline of young talent is drying up. And the FCA's response is to invest in spy tech for enforcement rather than making the profession worth entering.

You can't prosecute an algorithm. You can't strike off a chatbot. You can't hold a large language model personally liable under SM&CR. So who exactly benefits from a regulatory model built around punishing people, when the people are leaving and the machines are arriving?

We'd like your views:

  • Is the FCA's enforcement first model sustainable when the industry it regulates is shrinking and the unregulated AI advice market is exploding?
  • Should the FCA be actively making the human advisory profession more attractive, rather than more burdensome?
  • If AI financial advice sits outside the regulatory perimeter, who carries the liability when consumers lose money on hallucinated tax guidance?
  • At what point does heavy regulation of human advisers become a competitive advantage for unregulated AI — and is the FCA inadvertently accelerating that?
  • Should there be a mandatory human checkpoint for any AI generated financial recommendation that involves pensions, tax or debt?

 

5 responses from the Newspage community

Copy all

Star Quote
Copy

Colette Mason
AI Ethics Consultant at Clever Clogs AI
The FCA is buying surveillance tools to catch yesterday's criminals while tomorrow's consumer harm is already walking through the front door unchecked. Millions of people are taking pension and tax advice from rogue AI systems that sit outside the regulatory perimeter and answer to nobody. The regulator knows this. Its own Mills Review asks whether AI is already doing the job of regulated advisers without any of the obligations. But knowing isn't acting.

Financial regulation is built on a glacial political cycle: twice a decade, new MPs spawn new policies, consultations, and goals. The gap between regulation and technical capability is widening every week.

When the Mills Review reports back this summer, the AI tools it's examining will have been updated dozens of times. Regulation designed to govern people who can be struck off, fined, and jailed is being outrun by systems that can't be. The FCA is actively removing the advisers from the checkout process, leaving the public at risk.
Copy

Editor
Ben Perks
Managing Director at Orchard Financial Advisers
If they’re not careful, the FCA could regulate themselves into extinction.
Adviser numbers are lowering, the industry is less lucrative and attractive to youngsters and AI is coming, and fast.

With fewer people to regulate, the industry could be overrun by Robo-Advice. Surely it’s in the customers interest to have a Human Adviser.
A human is culpable, you can’t put cuffs ChatGPT.
A human is compassionate, we understand the quirky nuances that make advice differ from person to person.
But also, AI won’t need teams of people to regulate it. Jobs at the FCA are as under threat as the advisers within the industry.
It’s time for a change in approach. The FCA have done a great job in regulating the industry and forcing the crooks out. But now, after years of stringent regulation how many dodgy advisers remain?
Surely, now is a time for support, encouragement and enthusiasm towards advisers. Make the processes simpler, and give advisers the tools and scope to speak with more customers
Copy

Editor
Dil Gujral
Chief AI Trainer at AI Now Academy
Palantir should cause alarm bells.

It’s basically big brother, started by suing the US defence department.

No wonder it went pair shape.
Copy

Scott Taylor-Barr
Principal Adviser at Barnsdale Financial Management
Over regulation of legacy incumbents within a market, whilst simultaneously under regulating the newer digital entrants, is a mistake that has happened before, and one we are all still learning to live with. Social media platforms have grown into the main source of "news" for millions of people but do so with none of the safeguards that were built into traditional publishing and broadcast media, and this feels remarkably similar. As a profession advisers need to push the regulator to be robust with its treatment of AI models - it must be clear when a client is dealing with an Ai and not a human, there must be clear warnings and opt outs when a client is transacting on a non-advised and non-regulated (and therefore unprotected) basis. If AI’s are as good, or better, than people then they should be held to exactly the same standards and regulations too.
Copy

Rohit Parmar-Mistry
Founder at Pattrn Data
The FCA is building a stick and sawing off the branch it sits on.

If you push human advisers out with cost, paperwork and personal liability, consumers do not stop needing help. They go to whoever will answer at 11pm. Right now that is unregulated chatbots, influencers and scammers. When the advice is wrong, there is no firm to complain to, no PI insurer, no accountable Senior Manager.

The problem is not AI existing. It is AI doing regulated work without the bits that make advice safe: verification, suitability, audit trails and clear liability. In audits, the biggest red flag is when nobody can explain who owned a decision that used AI. Until the perimeter and liability model catch up, enforcement is theatre.

If the FCA wants safer outcomes, it needs both: fewer pointless burdens on humans, and hard rules for risk AI advice (pensions, tax, debt) including a mandatory human checkpoint.

Real question: who carries the can when an AI tool hallucinates guidance and people lose money?