China restricts OpenClaw AI use in government agencies over cyber risks

Chinese authorities have moved to restrict the use of OpenClaw AI applications within government agencies and state-run enterprises, citing potential security risks linked to the technology.

According to people familiar with the matter, employees at several government departments and state-owned companies have received notices warning them not to install OpenClaw software on office computers. In some cases, the restriction also extends to personal mobile phones connected to government networks.

The move highlights Beijing’s increasing caution toward emerging artificial intelligence tools that require access to large amounts of data and external internet services.

Government agencies warned against installing OpenClaw

Officials at several state-run banks and government institutions were reportedly told not to install OpenClaw on office systems.

Some notices instructed employees to inform supervisors if the software had already been installed so that security checks could be carried out. In certain cases, workers were asked to remove the application completely from devices used within official networks.

According to people familiar with the directives, some institutions also prohibited the installation of OpenClaw on personal phones that connect to government systems. In a few cases, restrictions were extended to the families of military personnel.

Authorities have not publicly announced a nationwide ban. Instead, some agencies require prior approval before the software can be used.

China’s Ministry of Industry and Information Technology and the State-owned Assets Supervision and Administration Commission did not immediately comment on the reports.

What is OpenClaw and why it is attracting attention

OpenClaw is an agentic AI platform that can perform tasks autonomously across digital tools. Unlike traditional chatbots that provide answers, AI agents like OpenClaw can take actions such as managing messages, organising files or making reservations.

The system can connect to platforms like messaging apps and workplace software. Users can give instructions, and the AI agent carries out tasks independently.

The platform quickly gained popularity among developers and technology enthusiasts after launching in late 2025. Many users demonstrated its ability to automate everyday tasks, which helped the technology gain attention globally.

Why China is concerned about the technology

Security experts say AI agents can require broad access to personal and corporate data. Because they interact with external services and process sensitive information, authorities worry they could expose systems to cyber threats.

Officials in China have long emphasised the importance of protecting data as part of national security strategy. The government maintains tight control over digital infrastructure through regulations and a heavily monitored internet system often referred to as the “Great Firewall”.

Analysts say tools that communicate externally or access sensitive datasets may raise concerns about data leaks or external interference.

Some cybersecurity researchers have warned that AI agents combine several risk factors: access to private data, the ability to communicate outside networks, and exposure to untrusted online content.

One expert described this combination as particularly risky if not carefully controlled.

Companies still exploring OpenClaw opportunities

Despite the restrictions, Chinese technology companies are continuing to experiment with the technology.

Major firms such as Tencent and JD.com have released their own versions of OpenClaw-based applications. Several local governments have also announced financial incentives to encourage startups to build services using the platform.

The growing interest has affected technology stocks as well. Shares linked to OpenClaw-related projects saw strong gains earlier, although some of those gains faded following reports of the government restrictions.

For example, AI developer MiniMax, which launched an agent system called MaxClaw earlier this year, has seen its market value rise sharply since listing.

The broader debate around AI agents

The restrictions reflect a wider global debate about how to regulate advanced AI systems.

Agentic AI tools can automate complex digital work, which could improve productivity. However, they also raise new questions around cybersecurity, privacy and accountability.

Governments around the world are studying how these systems interact with sensitive data and critical infrastructure.

China’s move suggests that authorities want tighter oversight before such tools become widely used in official networks.

For now, OpenClaw remains popular among developers and technology firms experimenting with AI automation. But the new restrictions show that security concerns could shape how quickly these tools are adopted across different sectors.