Alchemy and Visa launch AgentCard: AI agents get their own payment instruments
Crypto infrastructure platform Alchemy, together with payment giant Visa, has introduced AgentCard — a solution that empowers artificial intelligence to conduct online transactions on behalf of the user. This is not just a technical novelty, but a fundamental shift in how we perceive economic agents in the digital environment.
The integration with Visa Intelligent Commerce provides AI models with a full set of identification credentials: a virtual Visa card, an email address, a phone number, and a cryptocurrency wallet. Developers can connect the agent to pay for goods and services through a single API in less than a minute. The solution is already compatible with models from OpenAI and Anthropic, opening up broad opportunities for automation.
How It Works in Practice
AI agents gain the ability to book tickets, order groceries, and renew subscriptions — all without human involvement in the process. The system supports flexible spending limits, restrictions by store categories, and customizable budgets. By default, payments are processed through Visa tokens, preserving banking bonuses and credit lines for users. If the merchant accepts digital assets, AgentCard automatically switches to the crypto wallet.
"Every shift in technology has created new economic participants. AI agents are the next stage — they need access to the global economy," emphasized Alchemy CEO Nikhil Viswanathan. Visa, in turn, guarantees the security and scalability of such transactions through its proven infrastructure.
Context and Prospects
Previously, MetaMask announced a wallet for the era of autonomous AI, and Coinbase launched a service for connecting agents to user accounts for trading and payments. These events form a unified trend: the industry is preparing for the mass adoption of autonomous economic entities.
From my perspective, AgentCard is not just another product, but an important step toward creating a decentralized economy where AI agents become full-fledged market participants. However, the key challenges remain issues of security and control: how to ensure that the agent does not exceed limits or fall under the influence of malicious scenarios? The answers to these questions will determine how quickly this technology enters everyday life.