AI Shopping Gets a Human Check: World Launches Agent Verification Tool

As AI-powered shopping agents become more common, a new problem is emerging: how do websites know a real human is behind those automated purchases? World, a startup co-founded by Sam Altman, is stepping in with a solution.

Through its parent company Tools for Humanity, World has launched AgentKit, a beta tool designed to verify that AI agents making purchases are acting on behalf of real, unique humans. The system integrates biometric identity (via iris scans) with blockchain-based payment infrastructure to bring trust and accountability to the fast-growing world of “agentic commerce.”

Key Points

• Agentic commerce is rising fast: AI agents are increasingly being used to browse websites and make purchases automatically.
• New risks are emerging: This trend raises concerns about fraud, spam, and large-scale abuse online.
• AgentKit introduces verification: Websites can confirm that a real human is behind an AI agent’s decisions.
• World ID is central: Verification relies on a digital identity created from a user’s iris scan using World’s Orb device.
• Blockchain integration: AgentKit works with the x402 protocol, developed by Coinbase and Cloudflare, enabling automated transactions between systems.
• User approval is embedded: AI agents must be registered with a verified World ID, signaling that a human has authorized their actions.
• Still in beta: The tool is currently being tested with developers, with feedback expected to shape its future.
• Industry momentum: Major players like Amazon, Mastercard, and Google are already investing in agentic commerce infrastructure.

Key Quotes

• “AgentKit is built as a complementary extension to the x402 v2 protocol, in coordination with Coinbase.”
• “The integration is designed so that any website already using x402 can enable proof of unique human verification alongside (or instead of) micropayments.”
• “What the World ID badge tells you is that someone is a real and a unique human.”
• Comparing the system to legal delegation, Tiago Sada described it as giving AI agents a form of “power of attorney.”

Implications

• Trust layer for AI commerce: As AI agents take over purchasing tasks, verification systems like AgentKit could become essential infrastructure.
• Biometric identity goes mainstream: The use of iris scans for digital identity may see broader adoption if this model succeeds.
• Platform control shifts: Websites gain more control over which AI agents they trust, potentially shaping how agentic commerce evolves.
• World’s strategic positioning: By solving trust issues early, World is aiming to become the default verification provider in this emerging ecosystem.
• Balancing automation and accountability: The tool highlights a growing need to ensure that increased automation doesn’t come at the cost of security or reliability.

Source: https://techcrunch.com/2026/03/17/world-launches-tool-to-verify-humans-behind-ai-shopping-agents/