World links AI agents to human identities to curb bot abuse

The AgentKit framework allows users to connect AI tools to their World ID, ensuring that digital actions are tied to a unique human profile. By integrating with platforms such as Claude Code, Codex, Cursor, Hermes, and OpenClaw, the system enables users to delegate complex tasks while maintaining strict identity controls. World, backed by OpenAI CEO Sam Altman, intends for this infrastructure to help businesses enforce user-level rules, such as purchase limits, that are typically bypassed by large-scale automated scripts.

During the recent hat sale, the agents autonomously identified the product launch, navigated the storefront, and executed payments. Because the agents operated under verified World IDs, the system successfully enforced a one-person-per-item policy across participants in the U.S., Germany, Japan, and the U.K. This approach seeks to provide a middle ground for online services struggling to distinguish between legitimate delegation and coordinated bot behavior. As AI agents increasingly manage digital interactions, World’s model shifts the focus from securing accounts to verifying the human intent behind automated activity.