World Launches AgentKit Beta With Coinbase x402 Integration

World today unveiled AgentKit beta, a developer toolkit that enables AI agents to prove they are backed by a unique human via World ID, now integrated with the x402 protocol, an open standard created by Coinbase and Cloudflare.

—

The integration allows developers to build “human-backed agents” that interact with websites, APIs, and online services while maintaining privacy and trust.

Human-Backed Agents for the Agentic Web

AI agents are increasingly handling tasks once performed by humans—from booking reservations to comparing prices online. Analysts project agentic commerce could reach $3–5 trillion globally by 2030, with AI agents potentially accounting for up to 25% of U.S. e-commerce.

As autonomous agents become economic actors, platforms face a critical challenge: how to establish trust and accountability for automated interactions.

Loading tweet...

View Tweet

AgentKit solves this problem. Developers can allow verified users to delegate their World ID to AI agents, creating cryptographically verifiable proof that a real person stands behind each agent—without revealing their identity.

“Payments are the ‘how’ of agentic commerce, but identity is the ‘who.’ Integrating World ID with x402 gives developers a complete trust stack: agents can pay for services and prove a human backs them,” says Erik Reppel, Head of Engineering at Coinbase Developer Platform and Founder of x402.

“Proof of human addresses the gap platforms face in separating legitimate agents from bot swarms,” adds DC Builder, Research Engineer at World Foundation.

Overcoming the Bot Problem

Currently, most websites block automated traffic by default. While this prevents abuse, it also stops productive AI agents from performing legitimate tasks.

The x402 protocol introduced micropayments to limit excessive agent traffic. Since 2025, x402 has processed over 100 million payments across apps, APIs, and AI agents. But micropayments alone don’t establish uniqueness—a single person can operate thousands of agents paying small fees, creating challenges for platforms trying to verify legitimate users.

AgentKit adds a new layer: proof of unique human. Using World ID, users can verify their uniqueness without exposing personal data, and agents can inherit this verification. Platforms can now require payment, proof of unique human, or both before granting access.

Loading tweet...

View Tweet

How It Works

• Delegation: Verified World ID holders can assign their identity to multiple agents. Platforms can track how many unique humans are behind the agents.

• Access Control: Websites can limit or allocate resources based on unique humans rather than number of agents.

• Use Cases:

  • Reservation platforms can prevent scalpers from hoarding tables.

  • Ticketing platforms can ensure real fans, not bots, purchase tickets.

  • Free-trial services can allocate access per unique human.

• Privacy Features: Zero-knowledge proofs allow sharing only essential identity attributes (like age or country) without exposing full personal data.

World reports nearly 18 million verified humans in 160+ countries, providing a broad foundation for proof-of-human trust signals.

Loading tweet...

View Tweet

Developer Access

AgentKit beta is available now for developers with a verified World ID. Documentation and access details are provided on World’s developer portal. This beta is built on the existing World ID architecture, with a next-generation version planned to expand the capabilities of human-backed agents.

“AgentKit extends proof of unique human to the growing ecosystem of autonomous agents, creating trust and accountability while preserving privacy,” said the World team.

—

About World

World is building the largest inclusive network of real humans, offering proof of human, finance, and connection in the age of AI. Conceived by Sam Altman, Max Novendstern, and Alex Blania, World aims to enable every human to participate safely in digital networks.