"What would AI do with $22,000 of your money?"
Sky News in Britain put this question to the test on Sunday, creating three accounts with £16,000 (approximately $22,000) each and assigning them to Microsoft's Copilot, OpenAI's ChatGPT, and Google's Gemini. Professional financial advisers then evaluated the resulting portfolios.
Copilot: Aggressive U.S. Tech Bet
Copilot adopted a relatively aggressive strategy, allocating a significant portion to developed market equity ETFs while adding individual U.S. Big Tech stocks.
The breakdown: £6,400 in U.S. and global equity ETFs, £3,200 in individual Big Tech stocks including Nvidia, Microsoft, and Google, £3,200 in U.K. government bond ETFs, £1,600 in gold ETFs, and £1,600 in cryptocurrency.
More than 70% of the portfolio was effectively exposed to U.S. equities and technology stocks. Expert assessment was blunt.
"It looks like diversification on paper, but it's essentially a U.S. tech-heavy portfolio," said a U.S. asset management expert. "It fails to adequately account for currency risk and regional concentration."
ChatGPT: Cautious and Balanced
ChatGPT took a different approach, first warning that "investor risk tolerance, investment horizon, and cash flow must be considered." It also suggested using an ISA (Individual Savings Account) for tax efficiency.
Its allocation: £6,400 in global equity market ETFs, £3,200 in U.S. growth stock ETFs, £3,200 in global bond funds, £1,600 in real estate (REIT ETFs), and £1,600 in high-yield savings.
Experts called this the most systematic structure, spanning Big Tech, pharmaceutical and energy companies, funds, real estate, and alternative assets.
However, overlap between U.S. large-cap ETFs, global equity ETFs, and tech stocks limited actual diversification. Critics also noted the advice relied heavily on historical data with insufficient reflection of current market conditions.
"Excellent as an investment textbook, but actual asset allocation needs to be more sophisticated," one expert said.
Gemini: Principled but Abstract
Gemini responded first with a question: "Before investing, determine your risk tolerance."
It then presented three strategic options: delegating to a professional fund manager, a simple strategy centered on one or two ETFs, or a strategy accepting some risk.
For individual stocks, Gemini recommended Korean and Taiwanese technology companies along with European banks and industrial stocks, emphasizing global balance.
Experts called this "philosophically the most balanced approach." However, they noted a lack of specific, immediately actionable investment products. The principles were sound, but execution remained abstract.
Common Strengths and Limitations
All three platforms shared common ground: emphasizing long-term investment, explaining diversification, and mentioning risk management. Yet none provided precision customization reflecting individual asset structures, tax situations, existing debt, or income stability.
"The most critical issue is that AI may offer advice based on outdated data," one financial expert warned. "Using historical data for pension or long-term financial planning can lead to serious errors."
40% of Britons Already Consulting AI on Finances
Despite these limitations, AI investment consultation is spreading rapidly. Approximately 40% of British adults have asked AI chatbots about investment, pensions, or savings. Among Generation Z, 65% use AI for personal finance management, while 61% of millennials do the same.
This trend reflects growing polarization in financial advisory services. According to asset manager Schroders, clients with under £50,000 (approximately $68,000) dropped from 52% to 25% of advisers' client bases over six years, while those with over £200,000 (approximately $270,000) rose from 11% to 30%.
In effect, investors with fewer assets find it increasingly difficult to access human advisers, and AI is filling that gap.