More than 300 million online shoppers worldwide could regularly use AI agents for shopping and payments by 2030. This would represent approximately one in ten online shoppers, according to a new report by Mastercard, A Short History of the Future of Shopping and Payments.
The report was developed based on Mastercard’s research and predictions from four futurists from the United States, Europe and Asia: Magnus Lindkvist, Patrick Dixon, Katja Forbes and Theodora Lau. Among the changes they expect by the end of the decade are AI agents capable of negotiating prices, wearable devices that track how products perform after purchase and physical agents entering stores.
Recurring purchases such as groceries, medicines and subscriptions are among the first tasks consumers could entrust to agents. Lindkvist believes this division could also change the meanings we associate with purchasing.
“By 2030, purchasing and shopping will no longer mean the same thing. AI agents will handle routine purchases, while shopping will remain something we choose for discovery, inspiration and enjoyment,” said Lindkvist.
The change is already visible among younger consumers. Research conducted during June and July 2026 among 26,000 parents and teenagers aged 13 to 18 across 13 markets showed that teenagers use AI to support purchasing decisions approximately twice as often as their parents.
A fully automated assistant that recommends products, chooses among available options and completes a purchase according to predetermined rules would probably be used by 27 percent of teenagers and 16 percent of parents. Eighteen percent of teenagers use AI weekly to find the best price or discount, compared with ten percent of parents.
The results also reveal a different relationship with sources of recommendations. Almost one-third of teenagers, 31 percent, would trust a product recommendation from AI more than a recommendation from a friend. Twenty-three percent would favor artificial intelligence over advice from their parents. At the same time, 62 percent of young respondents expect this technology to significantly change how their generation shops.
For retailers, this development means adapting digital channels to a new type of customer. An AI agent can quickly analyze product data, verify claims, read reviews and complete a transaction. Information about prices, product features, loyalty programs and purchasing terms will therefore need to be structured so that both humans and machines can read and verify it.
Proving the claims brands use in their communications will be particularly important. If a product, for example, carries a sustainability label, the agent should be able to find and verify the certificate on which that claim is based. The quality and availability of data could therefore directly influence which products agents recommend or purchase.
The report identifies trust as one of the main conditions for the broader adoption of agentic commerce. Retailers, banks, fintech companies and payment service providers will have to define more precisely the agent’s identity, the scope of consent granted by the user, the method of transaction authorization and responsibility in the event of an error or misuse.
Lau predicts that at least three regulators from G20 member states will publish formal guidelines for the registration and oversight of AI agents by 2030. In such a system, the completed transaction will not be the only important element. Equally relevant will be the machine-readable authorization through which the user has previously specified what, when and under which conditions the agent is permitted to purchase.
Mastercard conducted the first live, end-to-end agentic payment transaction in Europe with its partners earlier this year. This example shows that the technology described in the report is already being tested in real payment systems, but its broader adoption will depend on security, transparency and the control retained by consumers.
