Consumers remain wary of AI agents handling full-service shopping

Personal AI agents like Meta’s Muse are being marketed as end-to-end shopping helpers, capable of researching, price-comparing, and checking out items. However, consumer sentiment remains cautious: a YouGov survey of over 3,000 adults in the United States and United Kingdom found that 53% would never let an AI make a purchase, and only 7% would allow it without prior approval. Accenture’s Consumer Pulse Research, covering more than 25,000 respondents worldwide, shows a similar pattern, with 74% trusting an AI more than a friend for recommendations but just 9% comfortable with fully autonomous buying.

Dan Coates of ACI Worldwide attributes the gap to trust issues and proposes an “agentic bill of rights” to set clear rules for agents, merchants, and banks. Payment networks like Mastercard and Visa are already developing such frameworks. Early user experiences have highlighted risks, such as a Muse-related incident where a YouTuber’s address was shared with a buyer on Facebook Marketplace. Executives believe trust will grow gradually as users start with familiar merchants and one-click ordering platforms.