Most Singaporeans ready to let AI agents shop for them with more safeguards: study
Singaporeans are among the most open to allowing an artificial intelligence agent to shop for them, but they also want the strictest guard rails, research shows. The latest agentic AI study from Global Payments released on Wednesday indicates the hoops agentic AI will need to jump through to be trusted to make purchases without human intervention. The study, which surveyed seven countries…
A recent study by Global Payments suggests Singaporeans are among the most open to allowing artificial intelligence agents to shop for them, but they demand strict safeguards. When surveyed across seven countries, including the US, China, and the United Kingdom, 59% of Singaporeans expressed willingness for an AI agent to make purchases, while 23% have already utilized such a service.
However, only 15% would allow an agent to shop, shortlist, and pay without human intervention. The research highlights the discrepancy between curiosity about AI for personal commerce, such as policy changes and travel planning, and the comfort in letting an agent do more than just browse and recommend.
Phil Pomford, Global Payments' executive lead for Enterprise APAC, noted that Singapore's eagerness for fast, verified payments is the same factor that prevents the adoption of AI agents. Consumers desire a confirmation step, certification, and proof of security before delegating the final purchase decision to an AI agent. The study found that 34% of Singaporeans would prefer a government-backed AI safety certification, the highest requirement among the surveyed countries.
Additionally, 41% requested a phone-push notification before an AI-driven purchase, compared to 36% globally. Security, particularly the protection of payment details, emerged as the primary concern driving this cautious approach.
Singaporeans cited several challenges in trusting AI agents for personal purchases. The most significant issues revolve around security, such as safeguarding payment information, alongside broader concerns about liability and fraud mitigation. These challenges are compounded by practical considerations, such as the potential for regret if an agent makes a purchase the consumer dislikes.
An earlier study by PwC and the Singapore FinTech Association revealed that 92% of Singaporeans were already using digital payment methods in 2025, with mobile wallets and QR codes being widely accepted. E-commerce sales in Singapore were projected to reach US$14 billion by 2027, nearly double the amount from the beginning of the decade.
However, the risk landscape is evolving, with scam schemes becoming more sophisticated and diverse in business models. Singapore is actively pursuing its role as a global AI hub, providing tax incentives for AI spending and investing over S$1 billion to fund public AI research over the next five years. An AI Council chaired by Prime Minister Lawrence Wong oversees strategic direction, yet Wong emphasized in February that AI must serve national interests and the well-being of Singapore's citizens.
Despite the potential for substantial AI involvement in personal finance, the study indicates Singaporeans would initially trust AI agents with relatively small transactions. On average, they were comfortable with an AI agent spending around S$49 on health and wellness or S$33 on a one-time digital purchase. Even in these cases, the respondents demanded the same level of scrutiny they would apply to wealth-management transactions.
Similarly, the survey showed that Singaporeans were comfortable with an AI agent conducting a modest S$400 money transfer or executing a trade valued just under S$1,200.
Written by urgent.news from South China Morning Post's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.