World is expanding access to AgentKit. This framework is used to bind AI agents to certified World IDs so that they can carry out tasks on the Internet on behalf of real users while retaining identity checks and user controls.
Proxy can bind real user identity
As AI agents become involved in more online operations, the platform and the business community face a more practical problem: How to distinguish between real users and large-scale automated robot networks. World offers the option of connecting proxy behaviour directly to their identity network.
According to the official introduction, users need to have validated World IDs and use World App to access supported AI tools. Current support tools include Claude Code, Codex, Cursor, Hermes and OpenClaw.
Upon completion of the connection, the user can bind the “real-person certificate” through the ToolRouter interface in World, generate the API key and access the AI proxy to the system within minutes. Thereafter, the agent will be able to represent the user in the website and application that supports AgentKit.
500 caps as demonstration cases
World recently demonstrated the practical use of the system in a limited-volume distribution round. The campaign launched 500 “Human in the Loop” hats, which are open only to certified World ID holders.
Officially, the 500 hats have been received by certified users in several countries and regions, including the United States, Germany, Japan and the United Kingdom. This case is mainly used to show that AI agents can complete real-world transactions with identity restrictions while reducing the risk of misuse by robots.
Aim at AI proxy online collaboration scene
According to World, with more service access, AgentKit, AI agents will not only handle information missions in the future, but also engage in broader online interaction and business activities. In this process, the key is not only automation, but also the ability of the agent to correspond to an independent and real person.
The project seeks to build a trust base for the “agent economy” by allowing AI agents to remain under user authorization and authentication while gaining greater autonomy. According to World, this is also part of its promotion of “human proof”, financial and connectivity services in the AI era.