New Delhi: Google has unveiled the Agent Payments Protocol (AP2), designed to securely initiate and process agent-led payments across platforms, according to an official statement.
Developed in collaboration with leading payments and technology companies, AP2 is an open protocol that supports a wide range of payment methods, including credit and debit cards, stablecoins, and real-time bank transfers. It aims to provide a consistent, secure, and scalable experience for both users and merchants while giving financial institutions the transparency needed to manage risk effectively.
At the heart of AP2 are Mandates—tamper-proof, cryptographically signed digital contracts that serve as verifiable proof of a user’s instructions. Signed with verifiable credentials (VCs), these mandates form the foundational evidence for every transaction. They address two main user-agent interactions: real-time purchases with human presence and delegated tasks without human presence.
AP2’s flexible design also supports both traditional and entirely new commercial models. To bolster support for the Web3 ecosystem, Google, in collaboration with Coinbase, Ethereum Foundation, and MetaMask, has launched the A2A×402 extension, a production-ready solution for agent-based crypto payments.
The company is already partnering with over 60 organizations—including Adyen, American Express, Ant International, Coinbase, Etsy, Mastercard, PayPal, and Salesforce—to shape the future of agentic payments. AP2 can also extend the Agent2Agent (A2A) Protocol and the Media Context Protocol (MCP). While A2A allows AI agents to communicate and collaborate, MCP provides a universal interface enabling AI models, particularly large language models (LLMs), to access external tools.
According to the statement, AP2 lays a trusted foundation for AI-driven commerce, establishing the core building blocks for secure transactions and creating opportunities for networks, issuers, merchants, technology providers, and end-users to innovate in areas such as seamless agent authorization and decentralized identity.
