By Eduardo Baptista
March 9 (Reuters) – Local governments in several Chinese tech and manufacturing hubs have announced measures to build an industry around OpenClaw, an AI agent seeing rapid adoption even as regulators warn of security risks linked to its access to personal data.
OpenClaw is an open-source AI assistant created by Austrian developer Peter Steinberger. It can perform tasks such as booking flights and organising email, effectively allowing a single individual to do the work of several people and spawning what have come to be known locally as “one-person companies”.
After appearing in November, the tool has become one of the fastest-growing projects in the history of GitHub, the world’s most widely adopted AI-powered developer platform. U.S. AI pioneer OpenAI last month hired Steinberger to build the next generation of AI agents.
OpenClaw has particularly taken off in China, where new technologies tend to be adopted very quickly. Tech giant Tencent hosted an OpenClaw setup session in the southern city of Shenzhen last week that drew children and retirees as well as developers.
SECURITY CONCERNS OVER OPENCLAW FLAGGED
Shenzhen’s Longgang district, which set up China’s first AI and robotics bureau last year, released draft measures on Saturday to build an OpenClaw-centred AI ecosystem and support “one-person companies”. It cited a recent central government report backing future industries such as humanoid robots.
High-tech development areas in Wuxi and Hefei in the country’s east, along with a city in the eastern manufacturing hub of Suzhou, all published similar draft measures focused on OpenClaw in recent days.
The growing popularity of OpenClaw and “one-person companies” have been highlighted at the ongoing National People’s Congress(NPC).
Zhang Xiaohong, a member of Jiangsu province’s NPC delegation and Communist party secretary of Soochow University, told Reuters that on-campus initiatives such as a competition to see which student can create the best “one-person company” were promoting practical AI skills.
But regulators and state media have also flagged security concerns around the agent, underscoring Beijing’s long-standing unease about cyber risks and data breaches.
The Wuxi measures stipulate that cloud platforms providing OpenClaw must ban access to sensitive data directories and should explore creating an AI compliance service center to focus on issues such as cross-border data transfers and IP protection.
SUBSIDIES FOR ‘ONE-PERSON COMPANIES’
