OpenClaw mania sweeps China as AI agent adoption overtakes the US

A viral wave of "agentic AI" is sweeping through China’s tech hubs, with the open-source framework OpenClaw achieving a level of mass-market penetration that has now surpassed adoption rates in the United States. The phenomenon, colloquially dubbed "raising a lobster" due to the project’s crustacean-themed branding, has transformed from a developer niche into a nationwide craze involving everyone from retired engineers to elementary school students.

Originally released in late 2025 by Austrian developer Peter Steinberger, OpenClaw has shattered GitHub records, amassing over 250,000 stars by mid-March 2026.

Unlike standard chatbots, OpenClaw functions as a "digital employee" capable of navigating operating systems, managing complex document workflows, and autonomously executing multi-step tasks like booking travel or coordinating across multiple software platforms.

The "lobster" economy and one-person companies

The "lobster-raising" craze has been fueled by a unique combination of corporate backing and aggressive local government subsidies. In cities like Shenzhen and Wuxi, authorities have rolled out "Lobster Ten Policies," offering up to 2 million yuan ($290,000) in grants for startups built on the OpenClaw framework.

These incentives are specifically targeting the "One-Person Company" (OPC) model, where individual entrepreneurs use fleets of OpenClaw agents to handle marketing, coding, and customer service.

Major Chinese tech giants have moved quickly to industrialize the hype:

• Tencent launched "WorkBuddy," a suite of OpenClaw-powered office tools integrated into its ubiquitous WeChat ecosystem.

• Baidu announced plans to embed OpenClaw directly into its mobile search app, potentially reaching 700 million users.

• Alibaba debuted "JVS Claw," an app designed to help non-technical users install and deploy their own agents in minutes.

  
  

A geographic flip in AI leadership

Industry analysts note that while the U.S. remains the leader in foundational AI models like GPT-4o, China has successfully "leapfrogged" the West in the deployment of autonomous agents. By pairing OpenClaw with domestic AI models that cost 60–80% less than Western alternatives, Chinese enterprises are automating white-collar workflows at a fraction of the cost.

"OpenClaw has tapped into a social sentiment regarding a new vision of productivity," said Li Zhi, head of the Intelligent Institute at Analysys International. "It is no longer about talking to the machine; it’s about the machine doing the work for you."

Security warnings and "regulatory friction"

The rapid rollout has not been without controversy. On March 11, Chinese authorities issued a series of alerts restricting state-run enterprises and government agencies from running OpenClaw apps on office hardware, citing significant cybersecurity risks. The framework’s requirement for high-level system permissions has raised fears of "agent-to-agent" data leakage and unauthorized access.

Despite these warnings, the "lobster" continues to grow. At recent "install fests" in Beijing and Shenzhen, thousands of citizens reportedly queued for hours to have volunteers help them set up their first autonomous agents, signaling that the era of the personal AI employee has officially arrived in the mainstream.