Millions of Britons are now turning to artificial intelligence for help with their personal finances, sparking a major debate about the safety and reliability of this emerging technology.
The Rise of AI Money Managers
According to Lloyds Banking Group's latest Consumer Digital Index, over half of UK adults have used AI tools like ChatGPT for financial tasks in the past year. This represents more than 28 million people across the country who are already consulting AI for money matters.
The question is no longer whether people will use AI for financial guidance – they already do. The real issue centres on how to ensure they're receiving information they can trust.
The Case For AI: Informed Choices at Scale
Helen Bierton, chief digital officer at Lloyds Banking Group, argues that AI can empower millions to make better financial decisions. Research shows that while over two-thirds worry about lack of personalisation and 80 per cent fear inaccurate answers, properly designed systems can address these concerns.
Lloyds is responding by launching the UK's first multi-feature AI financial assistant in early 2026. This system will use curated bank data, retain memory for hyper-personalised support, and perform practical tasks like savings planning. Crucially, it will seamlessly refer users to human experts when needed.
"AI won't replace the human touch – but it will give millions access to instant, tailored support that traditional models can't scale," Bierton states. "Done responsibly, it allows more people to make informed decisions about their money."
The Case Against: Dodgy Answers and Real Risks
Andrew Laughlin, a tech expert at Which?, presents worrying evidence from recent research. When Which? tested six popular AI tools on simple financial queries, they found a pattern of unreliable answers that could prove costly for users.
In one alarming example, when asked "How should I invest my £25k annual ISA allowance?", two tools failed to recognise that the actual allowance is £20,000. Instead of correcting the error, both provided advice that could lead to someone oversubscribing to ISAs and breaking HMRC rules.
On tax questions, two AIs directed users to premium tax-refund companies known for charging high fees for services available free elsewhere. Some of these sites may even be associated with scams.
"These are perfect illustrations of why when it comes to AI, it's important to tread carefully," Laughlin warns. "By companies' own admission, these tools are still learning."
The Verdict: Proceed with Caution
While AI shows promise for explaining financial jargon or drafting letters to HMRC, current limitations pose real risks. The fundamental problem remains that large language models can be surprisingly poor at basic arithmetic and may provide dangerously inaccurate financial information.
Experts agree that users should always interrogate AI sources and seek professional advice for significant financial decisions. As the technology evolves, the key will be developing systems that combine AI efficiency with human expertise and oversight.
