Affluent and high-net-worth investors (HNWIs) in Malaysia are adopting AI tools for finance and investment, but they still rely heavily on the critical oversight that comes from human professional judgement, according to an HSBC survey, released today. This is why more than half, or 58% still want human-AI synergy in making financial decisions.

A new survey of around 10,000 affluent and HNWIs across 10 markets, commissioned by HSBC and conducted by Ipsos, shows AI has become a widely used tool among these investors.

Malaysia ranked among the top three markets for AI adoption in finance at 85%, level with mainland China (85%) and trailing only India (86%). Aside from finance, these investors are also using AI for work and career (64%) and for personal development (64%).

Across the generations surveyed, Gen-Z (aged 21-29) and Millennials (aged 30-45) make up the heaviest users of AI for financial and investment decisions, at 86% and 89% respectively, followed by Generation X (aged 46 to 61) at 85% and Baby Boomers (aged 62-69) at 78%.

Despite the strong interest in AI for finance, these investors still rank financial professionals and institutions as the leading source of investment ideas (65%) and the most influential factor in investment decisions (39%), more than double the influence of AI tools (16%).

Linda Yip (pictured), country head of international wealth and premier banking, HSBC Malaysia, said: “Technology gives us speed, but human connection builds trust. The future of banking lies in a seamless partnership between AI-driven insights and human expertise. By pairing advanced analytics with human relationships, we aim to support our clients with the confidence to navigate, invest and pursue growth in an increasingly complex financial landscape”

When asked why investors turn to financial professionals and institutions when making financial decisions, 85% cite reassurance and 76% cite strategic expertise. This includes spotting mistakes in AI-generated data (31%), providing personalised interpretation of complex data (31%) and applying judgement and validation (30%).