Question 1

What is AgentKit and how does it work for AI agent verification?

Accepted Answer

AgentKit is a developer toolkit launched by Tools for Humanity in March 2026 that enables AI agents to cryptographically prove they are acting on behalf of a verified human. It uses World ID — the biometric identity credential from the Worldcoin ecosystem — to create a chain of trust between a human user, their AI agent, and the merchant or service the agent interacts with. When an AI agent attempts to make a purchase or access a service, AgentKit generates a zero-knowledge proof that confirms a real, unique human authorized the action, without revealing the human's identity or biometric data to the merchant. The system is designed to prevent unauthorized agent activity, fraud by rogue AI systems, and the proliferation of bot-driven transactions that lack human accountability.

Question 2

Why is human verification necessary for AI agents making purchases?

Accepted Answer

As AI agents increasingly browse the web, compare products, and execute purchases autonomously, merchants and payment processors face a fundamental trust problem: they cannot distinguish between an agent acting on legitimate human instructions and a rogue bot exploiting stolen credentials, executing unauthorized transactions, or gaming promotional systems. Traditional authentication methods like passwords and CAPTCHAs were designed to verify that a human is present — but in agentic commerce, the entire point is that a human is not present. A new verification layer is needed that confirms human authorization without requiring human presence at the point of transaction. Without this, merchants face escalating fraud risk, consumers lack recourse for unauthorized agent actions, and the entire agentic commerce ecosystem cannot scale beyond low-value transactions.

Question 3

How does AgentKit relate to Worldcoin and Tools for Humanity's broader strategy?

Accepted Answer

AgentKit is built on top of World ID, the proof-of-personhood credential that Tools for Humanity developed as part of the Worldcoin project. Worldcoin uses iris-scanning biometric hardware (the Orb) to create unique, privacy-preserving digital identities — over 12 million people have been verified as of early 2026. AgentKit extends this identity layer from human-to-service verification to human-to-agent-to-service verification, effectively making World ID the authentication backbone for autonomous AI commerce. The strategic logic is clear: if every AI agent transaction requires proof that a real human authorized it, and World ID becomes the dominant proof-of-personhood standard, then Tools for Humanity sits at the center of the agentic commerce trust layer — a position analogous to what certificate authorities became for HTTPS.

Question 4

What are the privacy and centralization risks of biometric identity for agentic commerce?

Accepted Answer

The primary concern is that biometric-based identity systems create a centralization chokepoint. If AgentKit or a similar system becomes the dominant verification layer for agentic commerce, a single entity effectively controls who can and cannot participate in autonomous AI transactions — a gatekeeping power with enormous commercial and civil liberties implications. Tools for Humanity uses zero-knowledge proofs to ensure that biometric data is not shared with merchants or agents, and the World ID system is designed to be privacy-preserving. However, critics argue that the initial biometric collection (iris scanning) is inherently invasive, that the company's privacy guarantees rely on trust in its cryptographic implementation, and that any system requiring physical biometric enrollment creates barriers to access. The risk of a biometric identity monopoly in agentic commerce mirrors concerns about credit bureau dominance in traditional finance — essential infrastructure controlled by a small number of private entities.

Question 5

How large is the market opportunity for agent identity and verification infrastructure?

Accepted Answer

The agent identity verification market is nascent but potentially massive. If AI agents mediate 8-12% of the $6.3 trillion global e-commerce market by 2028 — approximately $500-750 billion in transactions — and each transaction requires some form of human verification, the identity layer could extract 0.5-2% of transaction value as verification fees, representing a $2.5-15 billion annual revenue opportunity. This estimate does not include non-commerce agent verification use cases such as healthcare, financial services, government services, and enterprise procurement, which could multiply the market by 3-5x. For context, the digital identity verification market was valued at $10.9 billion in 2025 and is projected to reach $33 billion by 2030. Agent identity verification could represent the fastest-growing segment within that market.

Question 6

What alternatives to biometric verification exist for authenticating AI agents?

Accepted Answer

Several competing approaches are emerging. OAuth-based agent delegation models extend existing authentication frameworks to allow users to grant agents scoped permissions — similar to how users authorize third-party apps today. Blockchain-based decentralized identity (DID) systems like those from the W3C Verifiable Credentials working group enable agents to carry cryptographically signed credentials without a central authority. Hardware-bound authentication using device-level secure enclaves (Apple's Secure Enclave, Google's Titan chip) could tie agent authorization to a physical device the user controls. API key and token-based systems, already used by platforms like Shopify and Stripe, provide merchant-specific agent authentication. The question is whether any of these alternatives can provide the same level of assurance as biometric proof-of-personhood — particularly for high-value transactions where the stakes of unauthorized agent action are significant.