World, co-founded by Sam Altman, has introduced AgentKit, a tool designed to verify human identity behind AI shopping agents. This initiative aims to address rising concerns surrounding fraud and misuse as more consumers utilize AI agents for online purchases. The verification process hinges on World ID, derived from an iris scan using the World Orb device, which generates a unique encrypted digital code.
The introduction of AgentKit comes amid an increase in agentic commerce, where AI programs make purchases on behalf of users. While this trend offers convenience, it poses risks of fraud and internet abuse. AgentKit integrates with the blockchain-based x402 protocol, developed by Coinbase and Cloudflare, enabling seamless transactions without human intervention. Users register their AI agents with their World ID, which communicates human approval for purchases to merchants.
Tools for Humanity (TFH), the company behind World, states that AgentKit serves as a complementary extension to the x402 v2 protocol. According to the firm, this integration allows websites already employing x402 to authenticate unique human identities alongside traditional micropayments. TFH Chief Product Officer Tiago Sada described the function as akin to granting “power of attorney” to an agent, enhancing trust in transactions. “What the World ID badge tells you is that someone is a real and a unique human,” Sada explained, emphasizing that websites can still decide to block users they suspect of bad faith.
Currently, AgentKit is in beta, aimed at developers, with iterative feedback intended to improve the tool. To utilize this verification, consumers must secure their World ID through an Orb scan. E-commerce giants, including Amazon and Mastercard, along with Google’s own emerging protocol for agentic commerce, have adopted automated purchasing features, highlighting the trend’s growing acceptance.
