
What happened
Mastercard rolled out a payment option Thursday that gives an AI agent a virtual card to buy online without checking in each time. Cardholders can cap spending, restrict retailers, or require approval before checkout.
Why it matters
Letting a bot spend your money without a final click is exactly the step shoppers are least ready for, so the card networks' push to set the rules for bot-driven checkout may run ahead of consumer demand.
What to watch
Mastercard says its Verifiable Intent record could help resolve disputes, but it did not specify who would ultimately be responsible if an agent bought something outside the cardholder's instructions.
WHO IT HITSThe outcome lands hardest on payment-network product teams and checkout engineers, who must design authorization and dispute records for bot-driven purchases, and on online retailers whose checkout flow would change if agents buy on shoppers' behalf.
Summaries like this, in your inbox every morning.
Mastercard, Visa and Meta are all moving toward a checkout where an AI agent, not the shopper, completes the transaction. Mastercard's Thursday rollout gives an AI agent a virtual card with spending caps, retailer restrictions and optional approval steps, while Visa has announced Visa Intelligent Commerce and Meta's Muse can search and navigate checkout but still hands the final approval to the user. ACI Worldwide's Phil Bruno frames this as a "land grab for infrastructure standards," where setting the rules for agentic commerce keeps the commerce inside the winning network for decades.
The consumer signal in the same research points the other way. Just 7% of surveyed U.S. and U.K. fashion buyers said they would let an AI assistant purchase without approval under predefined conditions, and more than half said they were uncomfortable with AI buying on their behalf. What those shoppers value most is price-drop alerts and cross-retailer comparisons, each chosen by 35%, with personalized suggestions at 18%. Mastercard's response is Verifiable Intent, a record of who authorized the agent and what it was allowed to buy, which the company says could help resolve disputes; the company told Fortune that card protections in an agentic transaction would be the same as today's, but it did not say who would ultimately be responsible if an agent bought something outside those instructions.
The near-term test is whether clear liability and dispute records can move shoppers past the final click. That step is the hardest to hand over, as ACI Worldwide's Dan Coates put it: "We want to be the ones to make that final click."
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Applied Clinical Trials Online reports on lessons drawn from hundreds of clinical trials involving AI, coverin…

The RTX 5090, launched at $1,999, now costs over $6,000 at US retailers, with third-party listings above $9,00…

Palantir CEO Karp spoke bluntly in the AI risk debate, saying aloud what the article frames as the quiet part

Co-CEO Clay Magouyrk said Oracle last year hadn't found broad internal use for generative AI; after rolling ou…

Twist Bioscience signed an agreement to provide antibody characterization data and services to Eli Lilly's AI/…

At the AWS Global Meeting on September 17, Chief AI and Technology Officer Matt Wood named Trainium for comput…
