What happened
Visa and Mastercard are developing frameworks to support purchases initiated by AI agents, combining tokenized credentials, identity verification and permission-based transaction controls. The approach reframes tokenization, long used mainly for fraud prevention, as a way to validate that an agent has the authority to make a given purchase.
The networks aim to establish trust between consumers, merchants, issuers and autonomous software acting on a buyer's behalf. Both companies describe this as necessary infrastructure that must exist before agentic commerce can scale beyond experiments.
Why it matters
For AI agents to transact safely, there has to be a reliable way to confirm what they are permitted to buy and on whose behalf. Extending tokenization to encode authority and permissions gives the payment networks a role in that emerging layer of trust.
Without such controls, agent-driven purchases carry real risk of unauthorized or mistaken transactions, which would limit adoption.
MintedBrain take
Practitioners building agentic commerce should watch how these permission frameworks take shape, since they will define what agents can and cannot do at checkout. The involvement of the major card networks suggests standards, rather than one-off integrations, are where this is heading.
Discussion
Sign in to comment. Your account must be at least 1 day old.