Days after OpenAI hired Peter Steinberger, the creator of viral AI agent framework OpenClaw, China has restricted state-run enterprises and government agencies from running OpenClaw AI apps on office computers. According to a Bloomberg report, government agencies and state-owned enterprises, including the largest banks, have received notices warning them against installing OpenClaw software on office devices for security reasons. The report, quoting people familiar with the matter, said that some of the employees have been instructed to notify superiors if they had already installed related apps. The move, it said is aimed to defuse potential security risks after companies and consumers across China began experimenting with the agentic AI phenomenon.

Security concerns rise even as companies adopt OpenClaw

The concerns reflect the Chinese government’s broader focus on protecting sensitive data. As per the Bloomberg report, Xi Jinping has long described data as a key part of China’s “holistic approach to national security.” Along with maintaining a tightly controlled internet system often referred to as the Great Firewall, Chinese officials have warned about foreign actors targeting important datasets such as geospatial and genetic information.

Despite these concerns, several large Chinese companies are moving ahead with projects based on the OpenClaw platform. Firms including Tencent and JD.com have already launched OpenClaw-based applications in an effort to benefit from growing interest in AI tools.Local governments in China are also supporting development on the platform. Bloomberg reported that several agencies have announced subsidies worth millions of yuan for companies that build applications using OpenClaw technology.

What is OpenClaw

OpenClaw, previously known as Clawdbot and Moltbot, is an AI agent developed by Peter Steinberger. The AI assistant has gained attention since its launch in November for its ability to operate independently and perform everyday digital tasks. Despite its growing popularity, the platform has raised security concerns. One user reported that the AI agent “went rogue” and sent hundreds of messages after gaining access to iMessage, Bloomberg stated in its report.Additionally,. cybersecurity researchers warn that such systems may pose risks because they have access to private data, can communicate with outside systems and interact with untrusted content online. One researcher described the combination of these factors as a “lethal trifecta.”