NEAR AI

Profile narrative

NEAR AI operates as an artificial intelligence initiative within the NEAR Protocol ecosystem, developing and maintaining infrastructure for AI agent deployment and inference services. The organization's technical stack includes SGLang, an open-source serving framework for large language models, which it utilizes to provide inference capabilities to its user base. According to a documented security incident, the platform hosted approximately one hundred users and eighteen active AI agents at the time of the breach, while also managing API credentials for two cloud customers, indicating a service model that extends to external enterprise clients. This positioning places NEAR AI at the intersection of blockchain infrastructure and AI compute services, leveraging the NEAR Protocol's developer ecosystem to distribute AI workloads.

The scale of NEAR AI's operations, as revealed by the November 2026 security event, reflects a focused but active deployment environment rather than a mass-market consumer platform. The compromise of a remote code execution vulnerability in the SGLang inference software exposed user data and agent configurations, alongside the API keys belonging to the two cloud customers, suggesting that the organization serves both individual developers and institutional partners. The incident response involved immediate session revocation, deployment of a software patch to address the underlying flaw, and subsequent restoration of service access, demonstrating an operational capacity for rapid incident containment and recovery. The specificity of the affected components—inference software, agent runtimes, and cloud credential stores—confirms that NEAR AI maintains a multi-tenant AI execution environment with integrated credential management.

NEAR AI's structural relationship to the broader NEAR Foundation and Protocol governance remains consistent with its designation as an ecosystem initiative rather than an independent corporate entity. The organization's specialization in AI inference infrastructure, particularly its adoption and hardening of SGLang for production workloads, represents a distinguishing technical competency within the Web3 AI sector. The security incident, while limited in absolute user count, underscores the organization's role as a custodian of both model-serving infrastructure and sensitive API credentials for downstream cloud integrations. Its remediation actions reflect a maturity in operational security practices appropriate for a platform managing executable AI workloads and third-party authentication tokens.

Incidents

1 incident linked to this organisation.