China Overtakes US in OpenClaw AI Agent Adoption as Security Fears Mount

China Leads the OpenClaw Revolution

China has overtaken the United States in adoption of OpenClaw, the open-source AI agent framework that has become the fastest-growing open-source project in history with over 300,000 GitHub stars since its late 2025 launch. According to American cybersecurity firm SecurityScorecard, of the more than 142,000 publicly visible OpenClaw agents tracked worldwide, nearly half originated from China within roughly 100 days.

The Lobster Craze

The adoption frenzy has taken on a life of its own. On March 6, nearly a thousand people lined up outside Tencent headquarters in Shenzhen to have engineers install OpenClaw on their devices for free. Enthusiasts wear lobster-themed hats at meetups, and the phrase "raising the lobster" has entered popular slang. Tencent has released a suite of OpenClaw-based products branded "lobster special forces" integrated with WeChat, while Zhipu AI launched a local adaptation with over 50 pre-loaded skills.

Cost Is King

A key driver is economics. OpenClaw is model-agnostic, meaning it runs on AI models from any provider. Chinese open-source models from companies like MiniMax and Moonshot AI now offer competitive performance at a fraction of Western costs. Because AI agents consume far more computing tokens than simple chatbots, running continuously in the background, lower model costs create a self-reinforcing adoption loop. Local governments in Shenzhen and Hefei have proposed financing support of up to 10 million yuan for businesses developing OpenClaw applications.

Security Alarm Bells

The rapid spread has triggered a backlash. Chinese authorities have barred state-run enterprises and government agencies from installing OpenClaw on office systems. SecurityScorecard identified more than 135,000 publicly exposed instances across 82 countries, with over 50,000 exploitable through remote code execution vulnerabilities. Creator Peter Steinberger, who joined OpenAI in February and transitioned OpenClaw to an independent foundation, has acknowledged the security challenges of giving AI agents broad access to personal devices.