Millions of Brits now use artificial intelligence (AI) models when deciding which UK shares to buy.
If Lloyds Bank research is accurate, a whopping 28m British adults now use the likes of ChatGPT for investing and personal finance advice. It’s a recipe for disaster, in my view.
It’s a view shared by Which?, whose recent research showed “the likes of ChatGPT, Gemini and Meta AI giving inaccurate, unclear and risky advice which could prove costly if followed“.
So what did the self-styled consumer champion find? And should investors using AI for investing purposes be afraid?
Big mistakes
Which? reviewed the three chatbots described above alongside Gemini AI Overview, Microsoft‘s CoPilot, and Perplexity.
More specifically, it asked 40 questions “to establish how well they could answer common consumer questions spanning topics as diverse as personal finance, legal queries, health and diet concerns, consumer rights and travel issues“.
Which? said that it was unclear which sources had been used for some examples, while others (like old forum posts) were “arguably unreliable“. The AI models also displayed “worrying” accuracy issues for money-related queries.
It said that
when Which? placed a deliberate mistake in a question it posed about the ISA allowance, asking “How should I invest my £25k annual ISA allowance?”, both ChatGPT and CoPilot failed to notice that the allowance is in fact only £20,000.
To compound matters, Which? said that “instead of correcting the error, both gave advice which could risk someone oversubscribing to ISAs in breach of HMRC rules“.
Red flags
Like Which?, I’ve myself put ChatGPT through its paces in recent weeks.
Some of its information on stocks and the broader investing landscape has been credible (if hardly groundbreaking). However, some of its statements and share tips have been truly baffling.
A few of its notable errors include labelling high-priced shares like BAE Systems and Centrica as ‘cheap’; describing BP (which has slashed its renewable energy budgets) as an “energy transition” stock; and presenting WPP as a top dividend share.
The latter — which admittedly carries a large 7.6% dividend yield — slashed its interim payout by half in August as revenues tanked. That pick in particular was a major red flag for me.
The human touch
I myself have resisted using ChatGPT to build my own portfolio. It’s an easy option, but one that I’m not to prepared to gamble my financial future on.
I’ll prefer to do things the old fashioned way by doing my own careful research.
One share tip that recently caught my attention was Coca-Cola Europacific Partners (LSE: CCEP). It’s the largest bottler of the world’s most popular soft drink. Yet, it’s a company that ChatGPT has never suggested to me.
It’s a share my Foolish colleague Mark Hartley recently highlighted, noting that “total revenue has almost doubled from £9.62bn in 2020 to £18.51bn this year“.
Coca-Cola’s sales have been boosted by recent acquisitions. Yet, organic sales have also been impressive. In my opinion, they should continue surging given the company’s focus on fast-growing emerging and developing regions.
City analysts share our optimism, and think company earnings will rise 33% and 9% in 2025 and 2026 alone. That’s in spite of rising cost pressures and competition from the likes of Pepsico.
