Several Chinese government offices and state-owned firms have recently instructed staff not to install the artificial intelligence agent OpenClaw on work devices, citing security risks, according to individuals familiar with the advisories. The move comes as investors follow both the continued rise of enterprise AI adoption and the growing scrutiny over digital security in China.
OpenClaw, previously known as Clawdbot and Moltbot, launched last November and has rapidly gained traction for automating digital tasks such as clearing emails, booking restaurants, and checking in for flights. The software connects with platforms like WhatsApp and Slack, allowing users to manage tasks autonomously through messaging apps.
OpenClaw was developed and uploaded to GitHub by an Austrian developer, Peter Steinberger, in November, 2025. Later, he was hired by OpenAI in January, 2026.
Chinese tech developers, prominent AI firms, and local governments in major technology and manufacturing regions have actively promoted and deployed the tool over the past month.
Despite this widespread adoption, notices circulated in recent days directing employees at government organizations and major state-owned enterprises—including leading banks—not to install OpenClaw on office hardware.
Some instructions also ask staff to disclose if they have already installed associated applications for potential removal following security screening, sources said on the condition of anonymity. In certain cases, restrictions extend to personal phones connected to work networks, and, for some, even to family members of military personnel.
Authorities’ actions reflect Beijing’s efforts to balance a national “AI plus” strategy aimed at integrating artificial intelligence across the economy with persistent concerns over data protection. OpenClaw’s operational model, which demands extensive access to individual user data and facilitates external communication, has amplified official worries about exposure to cyber threats, especially as geopolitical tensions rise.
Within some government offices, blanket bans have not been imposed, but staff have been cautioned about the cybersecurity implications of using OpenClaw. The breadth of these warnings is not fully known, nor is the potential impact on local authority plans, some of which involve substantial subsidies for companies innovating with OpenClaw in line with Beijing’s policy framework.
President Xi Jinping has repeatedly placed data control and security at the center of national policy, citing external threats to sensitive datasets as a prominent security risk.
Although these measures reflect concern about cybersecurity vulnerabilities, many large Chinese technology companies, among them Tencent Holdings and JD.com, continue to integrate OpenClaw into their offerings to capitalize on the current surge of interest in agentic AI platforms.
