The framework links AI agents to a verified World ID, allowing websites and applications to determine whether an agent represents a unique human user.
World, a crypto and digital identity project built to prove real human identity online backed by OpenAI CEO Sam Altman, is broadening access to AgentKit, a framework designed to connect AI agents with verified human identities through World ID.
The platform aims to help businesses distinguish between AI agents acting on behalf of real individuals and automated bot networks as autonomous software becomes more prevalent online.
AgentKit enables users to delegate tasks to AI agents while preserving identity verification and user controls. By linking an AI agent to a verified World ID, websites and applications can confirm that the agent represents a unique human rather than a mass-operated automation system, the team stated.
According to World, individuals must have a verified World ID and World App account, along with a supported AI agent such as Claude Code, Codex, Cursor, Hermes, or OpenClaw, to use the service. Through World’s ToolRouter interface, users can generate credentials and connect their agents to supported services.
The technology was recently showcased through a limited-edition sale of 500 “Human in the Loop” hats. During the event, AI agents completed the entire purchasing process on behalf of users, including discovering the product release, verifying eligibility, navigating the online storefront, and executing transactions. Identity verification ensured that purchase limits were enforced on a one-per-person basis.
All hats were successfully claimed by verified users across countries including the US, Germany, Japan, and the UK. World said the demonstration illustrated how businesses can safely allow AI agents to perform online actions while reducing abuse from bots and automated farming networks.
Disclosure: This article was edited by Vivian Nguyen. For more information on how we create and review content, see our