Abstract

x402 is an emerging payment protocol for Web APIs and autonomous AI agents that uses blockchain for settlement. Because it relies on third-party "facilitators" to manage payment verification and on-chain settlement, these intermediaries represent a centralized point of trust and potential failure. This paper presents the first systematic study of security risks in real-world x402 deployments, identifying critical vulnerabilities in facilitator implementations that can lead to financial loss, asset theft, and service disruption.