Paradigm, a US-based venture firm with USD 12.7 billion in assets under management, and Stripe, the payments company valued at USD 159 billion, have launched Tempo, a payments-focused blockchain, alongside the Machine Payments Protocol (MPP), an open payment standard designed to enable AI agents to make purchases autonomously without human involvement in the payment flow.
First announced in September 2025, MPP provides a standardised framework for AI agents and service providers to exchange payment instructions directly, removing the complexity linked to account creation, card entry, billing forms, and manual transaction verification that the current infrastructure requires. A practical example would be a coding assistant automatically requesting additional computing resources from a cloud provider, receiving a quoted price, paying from its wallet, and unlocking the service, all without human intervention.
Multi-rail support and ecosystem partners
A key distinction of MPP from earlier protocols is its positioning as a multi-payment-rail framework from the outset. Visa has extended MPP to enable card payments across its network, while Lightspark has adapted the protocol for bitcoin payments over the Lightning Network. This means AI agents using MPP can transact via stablecoins, traditional cards, or bitcoin, depending on the context and counterparty.
MPP has already been integrated into more than 100 services spanning model providers, developer tools, compute platforms, and data vendors. Partners in the Tempo ecosystem include Anthropic, DoorDash, Mastercard, Nubank, OpenAI, Ramp, Revolut, Shopify, and Standard Chartered.
Furthermore, the protocol enters a market where a comparable standard already exists. Coinbase and Cloudflare launched a similar protocol, x402, in May 2025, which has processed approximately USD 34 million in payments to date, predominantly in stablecoins. Stripe and Paradigm are positioning MPP as a broader framework by incorporating card and bitcoin rails alongside stablecoin settlement.
The commercial rationale reflects expectations around the growth of agentic commerce. McKinsey estimates that by 2030, AI agents could mediate between USD 3 trillion and USD 5 trillion in transactions globally, characterised primarily by smaller, faster, and more frequent payments rather than large, occasional transactions.
