China is cracking down on open-source AI agents
The Chinese government has issued a strict warning to its agencies and state-owned enterprises not to install the artificial intelligence agent OpenClaw on office computers due to security concerns, according to recent reports by Reuters and Bloomberg.
OpenClaw is an open-source software capable of performing a wide range of tasks autonomously with minimal human input, extending beyond the typical research and question-answering functions associated with AI chatbots.
During the past month, the tool has been widely adopted and actively promoted by Chinese technology developers, leading AI companies, and several municipal governments located in the country’s major technology and manufacturing centers.
At the same time, regulators from the central government and state media outlets have warned that OpenClaw could unintentionally expose, remove, or misuse user data once installed and granted system permissions to operate on a device.
The restrictions suggest that while Beijing is encouraging the expansion of its “AI plus” strategy, which aims to integrate artificial intelligence across the economy to drive innovation-led growth, authorities are also cautious about potential cybersecurity and data protection risks amid rising geopolitical tensions.
It remains unclear how widely the restrictions are being enforced or whether they will influence policies at the local government level. Some municipalities currently offer subsidies worth millions of dollars to companies developing projects with OpenClaw as part of local implementations of Beijing’s national “AI plus” initiative.
Last week, a research institute affiliated with Shenzhen’s municipal health commission hosted an OpenClaw training program that drew thousands of participants and promoted its use in healthcare as part of the broader “AI plus” campaign.
OpenClaw was created by Austrian developer Peter Steinberger and released on GitHub last November. Steinberger joined OpenAI last month.
