By 2030, AI agents will reach significant mainstream adoption, according to a new report by Mastercard.
More than one in ten online shoppers will routinely use AI agents to purchase products on their behalf, a trend that’s already being embraced rapidly by younger generations. At the same time, AI agents will transform the payments landscape by automatically optimizing consumers’ wallets and autonomously managing personal finances.
However, as agentic AI becomes prevalent, establishing governance and compliance-control frameworks for these agents will become essential. These frameworks will ensure the verification of identity, authority, intent, and accountability, mitigating risks relating to fraud, privacy, and regulation.
These are some of the key predictions made in the Mastercard report. Drawing on Mastercard research and insights from world-leading futurists from the US, Europe and Asia, the report offers predictions about the future of shopping, retail and payments in the era of AI agents.
Agentic commerce adoption will vary globally
Patrick Dixon, a futurist who has advised over 400 major corporations, including Google, IBM, Bank of America and the World Bank on global trends, strategy and risk over more than three decades, predicts that by 2030, 300 million shoppers will use AI agents on a regular basis for their convenience. However, the adoption of agentic commerce will vary across different countries, reflecting the levels of trust in e-commerce and differences in online access.
In particular, significant differences will arise in Europe owing to the varied levels of e-commerce adoption across the region. For example, in the UK, 28% of all retail sales are already online, compared to only 14% in France or 16% in Germany.
This disparity is also evident in Mastercard’s own research findings, which found that 22% of UK teens expect shopping to be AI-led within five years, while in France it’s only 11%. Hence, Dixon predicts that agentic commerce will accelerate twice as quickly in the UK compared to France over the next two to three years.
Similarly, in some emerging markets like Colombia and South Africa, online commerce hasn’t taken off as strongly as in nations like the US. Currently, only 7% of retail in Colombia, and 10% in South Africa, is online, suggesting that in these countries, agentic commerce will emerge much more slowly.
Younger consumers are already embracing AI
AI is already being embraced for shopping, particularly among teenagers. A survey conducted in June and July 2026, commissioned by Mastercard and involving 26,000 parents and teenagers aged 13-18 across 13 countries, found that today’s teens are using AI to assist in shopping decisions at approximately twice the rate of their parents. Moreover, they are adopting AI shopping tools at a pace that surpasses the rate at which their parents have adopted similar technologies.
27% of teens expressed a likelihood of using a fully AI-run shopping assistant to recommend products, choose between options and complete purchases based on rules they set, compared with 16% of parents. Additionally, teens are twice as likely as their parents to be using AI weekly to find the best price or discount at 18% versus 10%.
Furthermore, there is a generational shift in trust towards AI. A significant portion of teens, 31%, indicated that they would trust an AI’s product recommendations over those of a friend. 23% of teens now trust AI more than their own parents’ advice.
Payment intent becomes a financial primitive
Theodora Lau, a prominent fintech expert, public speaker and the founder of Unconventional Ventures, anticipates that by 2030, verifiable consumer intent will emerge as a crucial foundational element of financial services.
Verifiable intent, a framework introduced in March 2026 by Mastercard and co-developed with Google, establishes a seamless connection between identity, intent, and action. It compiles these elements into a single, privacy-preserving record that captures the cardholder’s precise authorization and instructions. This creates a comprehensive shared audit trail that allows consumers, merchants, and issuers to confidently rely on if a dispute arises.
Lau predicts that financial institutions that build credit products and savings recommendations on top of verifiable intent will be able to deliver better products and services compared to those relying on historical transaction proxies and behavioral inferences alone.
AI agents will continuously re-optimize consumers’ wallets
By 2030, Lau anticipates that AI agents will optimize wallets for the objectives, preference and constraints defined by consumers, which may include economic, behavioral and personal factors. These agents will learn how the consumer wants their money managed and continue to refine their choices against their feedback over time, becoming closer to a financial chief of staff for the consumer.
Cryptocurrency startups are making significant strides in this direction. In February, Coinbase introduced Agentic Wallets, a wallet infrastructure built specifically for agents. This infrastructure enables agents to autonomously spend, earn, and trade while maintaining enterprise-grade security and programmable guardrails.
Similarly, MetaMask, a leading self-custodial wallet developed by Consensys, launched in June 2026 the MetaMask Agent Wallet, a self-custodial agent wallet that lets AI agents trade autonomously across the Ethereum ecosystem.
Traders just need to connect their agent framework and choose an operating mode. The Guard Mode is the default setup with daily spend limits, allowlisted protocols, and human approval via two-factor authentication (2Fa) on anything outside the policy set by each user. The Beast Mode is opt-in for traders and developers who want fewer interruptions, with 2FA still triggers on transactions flagged as malicious, but not on every policy edge case. Once deployed, the agent operates autonomously inside the trader’s rules.
“Know-your-agent” becomes a compliance advantage
As AI agents operate across multiple merchants and initiate transactions at machine speed, the traditional know-your-customer (KYC) and multi-factor authentication frameworks designed for human actors will need to evolve. Furthermore, Lau expects at least three G20 financial regulators will issue formal guidance on how AI agents must be registered, authorized, monitored, and held accountable by 2030.
In 2026, several industry initiatives emerged, including frameworks from MetaComp, a Singapore-based fintech platform, and Experian, a global data broker and consumer credit reporting company.
MetaComp’s StableX KYA Framework, launched in April 2026, establishes how AI agents are identified, authorized, monitored and held accountable to provide financial services in payments, compliance, and wealth management.
Experian’s Agent Trust, meanwhile, introduces a KYA framework that connects identity, intent and risk. It verifies and confirms the individual and payment method, issuing a KYA trust token. Subsequently, it establishes a connection between the consumer, device and AI agent, creating a binding relationship between human and agent. Finally, it executes the tokenized payment.
Featured image: Edited by Fintech News Singapore, based on image by tete_escape via Magnific
