China bans OpenClaw AI use in banks and state agencies over security fears

Chinese authorities have restricted state agencies and banks from running the OpenClaw app on office computers over the growing security concerns posed by the agentic AI system.

The decision comes after China, earlier this week, witnessed a massive boom in the adoption of OpenClaw. The users have rushed to the major Chinese tech firms to install this agentic AI software, calling this trend “raising a red lobster.”

Companies including Tancent, Baidu, and Alibaba have launched on-ramp services to make the software easier to install.

In the midst of growing OpenClaw hype, the state-owned enterprises and government agencies have been restricted from installing the software for security reasons, such as data breaches and exploitation, as reported by Bloomberg.

According to the sources privy to matter, the companies which have already installed the software, are given instructions to uninstall OpenClaw. One source said, the ban was extended to military personnel and their families.

While the platform has triggered a massive stock market rally and a wave of corporate adoption from giants like Tencent and MiniMax, the Chinese government is wary of the "lethal trifecta" of risks: broad access to private data, the ability to communicate externally, and exposure to untrusted content.

There are also reports where agentic AI systems went rogue and spammed hundreds of messages through iMessage.

However, some contradictions exist when it comes to local government actions.

In a paradoxical move, local governments in Wuxi and Shenzhen are offering millions of yuan in subsidies to startups developing on OpenClaw.

For the first time, AI agents were included in the Chinese government’s annual work report, with Premier Li Qiang calling for their "large-scale commercial application.

OpenClaw, formerly known as Clawdbot or Moltbot, is an autonomous AI agent capable of managing calendars, emails, travel check-ins, and restaurant reservations.