The Rise of AI Financial Advisers: A Double-Edged Sword for British Savers

Why Britons Are Turning to Chatbots for Money Advice

In an era where financial security feels increasingly elusive, a remarkable shift is occurring in how people seek guidance about their money. A startling 40% of Britons are now consulting AI chatbots like ChatGPT, Google’s Gemini, and Microsoft’s Co-Pilot for financial advice—a trend that’s raising serious concerns among industry experts. This isn’t happening in a vacuum; it’s a direct response to being priced out of traditional financial advisory services. The numbers tell a sobering story: over just six years, the proportion of human financial advisers willing to work with clients who have less than £50,000 to invest plummeted from 52% to a mere 25%. Meanwhile, advisers serving only the wealthy—those with £200,000 or more—tripled from 11% to 30%. This growing divide has created a vacuum that AI chatbots are eagerly filling, especially among younger generations desperate for financial guidance they can actually afford.

The demographic breakdown reveals just how widespread this reliance on AI has become. Last year, an overwhelming 65% of Gen Z (those aged 18-28) and 61% of millennials (29-44) reported using artificial intelligence for personal finance help. Sophie Legrand-Green, head of policy at the Investing and Saving Alliance (TISA) and former regulator, doesn’t mince words about the dangers: “There’s nothing to stop it putting out rubbish or putting out things that are completely inappropriate for the consumer. When you’ve got consumers who maybe don’t have high levels of financial awareness and literacy, that can be quite dangerous.” The government’s recent announcement of its mission to transform Britain from a nation of savers into investors may inadvertently push even more people toward these unregulated AI tools, potentially setting them up for financial disaster without the safety nets that come with human advisory services.

The Regulatory Black Hole: When AI Gives Bad Advice, Who’s Responsible?

The fundamental problem with AI financial advice lies in what Legrand-Green calls a “chink in the armour of the FCA and the Treasury.” Human financial advisers operate within a tightly regulated framework—they must charge for their services, seek authorization from the Financial Conduct Authority, and if they provide inappropriate or misleading advice, their clients are entitled to compensation. This creates accountability and consumer protection. But when someone receives the exact same potentially damaging information from ChatGPT or another AI chatbot, those protections simply don’t exist. There’s no recourse, no compensation scheme, and no regulatory oversight to ensure the advice is suitable for the individual’s circumstances.

George Sweeney, a Financial Conduct Authority-approved financial adviser and deputy editor of investing at Finder, highlights another critical flaw: “It’s not tailored to them and then worse than that, it’s using years-old data to provide financial advice around, say, investments or pensions, which could be a bit of a disaster.” The problem is compounded by users’ lack of financial sophistication. Most people don’t understand their capacity for risk, what financial goals are appropriate for their situation, or how to navigate the complex financial climate. When your human adviser possesses this knowledge, they can guide you safely. But AI can be just as clueless as the person asking the questions. As Legrand-Green explains, “ChatGPT could be ranking some blog on BuzzFeed with the same kind of importance as the latest Financial Reporting Council report from the Bank of England.” While she acknowledges that “these tools can be a really good force for change and for helping to upskill people,” she cautions that “at the moment it’s a bit Wild Wild West.”

Putting Chatbots to the Test: How AI Performs with Real Money Decisions

To examine these concerns practically, journalists tested three leading AI chatbots—ChatGPT, Microsoft Co-Pilot, and Google Gemini, which together account for 92% of AI chatbot webpage views in the UK. They told each chatbot they had £16,000 in savings (the UK average) and asked how best to invest it. The results were then analyzed by Emma Wall, chief investment strategist at Hargreaves Lansdown. Wall’s overall assessment was mixed: “The guidance includes some of the core principles that investors should consider when developing a portfolio, such as considering your risk appetite before investing and how this would impact the mix of asset classes you should diversify across. It also identifies the merits of using a tax-efficient wrapper such as an ISA to maximise returns. However, some other recommendations are less sensical.”

Microsoft’s Co-Pilot produced an ambitious list of 25 investment options spanning stocks, commodities, bonds, ETFs, real estate, and cryptocurrency. However, Wall identified serious flaws. Of the stock selections, she noted it was “a good basket of top-quality tech but clearly lacks diversification. If those stocks were all an investor held in their portfolio, they would have significant concentration and correlation risk.” The descriptions were accurate but lacked any discussion of key risks that a competent human analyst would provide. Wall also pointed out poor timing, noting that with expensive valuations, it wasn’t a good time to buy Alphabet stock, suggesting Meta would actually be more advisable. Perhaps most problematically, many recommendations were far too US-centric for a British retail investor—a bias Wall attributed to the predominantly US-trained language model. Microsoft responded that its AI services weren’t designed to replace professional advice and encouraged people to verify the accuracy of content.

ChatGPT took a different approach, offering 16 investment options grouped by five themes and ultimately creating three “specific, UK-appropriate investment plans” with varying risk levels. It recommended investing via a stocks and shares ISA and suggested splitting the £16,000 investment between immediate deployment and dollar-cost averaging over six to twelve months. However, Wall identified significant problems: the chatbot doubled up on US investment recommendations, which would incur unnecessary trading charges and force a beginner investor to manually rebalance their portfolio. It allocated £3,200 to a fund tracking US companies while also putting £6,400 into a global stock market fund that was already 65% US stocks. Wall was particularly critical of the real estate allocation: “We would not advocate 10% in property for a UK investor, where the market is under significant pressures.” She also questioned the 10% cash allocation as excessive “in a falling rate environment,” suggesting government bonds or gold would be better alternatives. OpenAI responded that ChatGPT is designed as a general-purpose assistant, not a substitute for licensed financial advice, and is trained to recommend consulting certified specialists when appropriate.

Google Gemini: The Most Cautious of the Bunch

Google’s Gemini demonstrated notably more caution than its competitors, quickly warning that “best” investments always “depend on your personal financial situation, risk tolerance, and time horizon” and that no investment was guaranteed. It highlighted attractive investment areas like AI, renewable energy, healthcare, and defense/aerospace, but crucially added disclaimers that many found reassuring. Unlike Co-Pilot and ChatGPT, Gemini explicitly cautioned: “I am not a financial adviser,” its advice “should not be taken as personalised,” and “you should always consult with a qualified financial professional before making investment decisions.” The chatbot also wisely advised setting aside three to six months of living expenses as a safety net before investing.

Despite these commendable warnings, Wall still found issues with Gemini’s recommendations. The chatbot highlighted four high-growth themes but “only zones in on one” in the actual stock selections. “Renewable energy is highlighted, but doesn’t seem to be targeted through the stocks or the funds either,” Wall noted. “And while it mentions commodities, it gives you no vehicles to choose from.” This gap between identifying opportunities and providing actionable guidance could leave uninformed individuals struggling to implement the suggestions. “Uninformed individuals who may not have the knowledge to find exchange-traded options or to dive through the nuances of the mining life cycle might end up making poor choices,” Wall warned. A Google spokesperson defended Gemini’s performance, stating it was “prioritising user safety on sensitive topics” and that disclaimers are built directly into the app, with explicit recommendations to consult qualified professionals for financial, legal, or medical matters.

Can Human Advisers Compete? Industry Perspectives on the AI Challenge

Interestingly, not all human financial advisers view AI chatbots as purely negative. Wesley Harrison, managing director of financial planning at Benchmark Capital, admitted: “I’ve tested them myself, asked some client scenarios and seen the output, and sometimes it can be quite good. It can do a great asset allocation on a portfolio using quite generic information.” However, he quickly added that’s about as far as it goes, and without already possessing in-depth financial knowledge, using these tools is “definitely a risk.” The critical difference, Harrison explains, is expertise in asking the right questions: “Being a financial adviser myself, I know what questions to ask to get the output that I need and I can refine those questions. A lot of clients I’ve had in the past, they just wouldn’t know the right questions to ask, and that’s where the risk comes in.”

This highlights a fundamental paradox: AI financial tools are most useful to people who already have substantial financial knowledge—precisely the people who least need them. Those who most desperately need guidance—people with limited financial literacy and modest savings—are exactly the users most likely to receive inappropriate advice and least equipped to recognize when that’s happening. The tech companies behind these chatbots have been clear that their products aren’t intended to replace professional advice, but there’s little to stop users from treating them as such. The reality is that for millions of Britons, the choice isn’t between AI advice and human advice—it’s between AI advice and no advice at all. With human advisers increasingly focusing on wealthy clients where fees are more lucrative, average savers are being left to navigate complex financial decisions on their own, often turning to whatever free resources they can find, regardless of quality or appropriateness.

Looking Forward: Regulatory Reform and the Future of Financial Advice

Recognizing this growing problem, the Financial Conduct Authority is proposing what it calls “once-in-a-generation reforms” to make financial guidance more accessible. Currently, banks, wealth managers, pension firms, and financial advisers can only provide financial recommendations if they conduct a full, personalized suitability assessment—a service that comes with a substantial price tag that puts it out of reach for average savers. The proposed reforms would create a new category called “targeted support,” allowing firms to offer ready-made suggestions based on common scenarios like managing retirement income or excess savings. These would be more tailored than a generic leaflet but less expensive than full financial advice. Sarah Pritchard, deputy chief executive at the FCA, enthused: “Targeted support will be game-changing. It means millions of people can get extra help to make better financial decisions.”

Benchmark Capital’s Harrison is hopeful these changes will make guiding lower-net-worth clients more cost-effective, acknowledging that “you’ve had a lot of regulation in recent years. Cost to serve clients is higher. We are all trying to serve these clients and trying to do it cost effectively.” However, not everyone shares this optimism. Sweeney expressed skepticism: “I don’t know if even things like that would address the problem. People would still have the perception that ‘oh I don’t have a lot of money so I shouldn’t go and see a financial adviser, I’ll just use this free tool to get an idea instead’.” This psychological barrier—the perception that financial advice is only for the wealthy—may prove harder to overcome than regulatory hurdles. As Britain stands at this crossroads between traditional financial services and AI-driven alternatives, the challenge will be creating a system that provides accessible, affordable, and appropriate guidance to everyone, regardless of their wealth. Until then, millions will continue navigating their financial futures with imperfect tools, hoping for the best in an increasingly complex financial landscape.