AI bot hype in China: Warning against using OpenClaw in authorities and banks
There is enormous hype around the powerful AI bot OpenClaw in China. A warning from the cybersecurity center suggests where it has all been installed.
(Image: muhammadtoqeer / Shutterstock.com)
China's cybersecurity authority CNCERT/CC has warned against the uncritical use of the enormously popular AI bot OpenClaw and advised authorities, state-owned enterprises, and banks not to install the technology on work devices. This is according to a notice from the National Computer Network Emergency Response Technical Team/Coordination Center of China on WeChat and a report by Bloomberg. Those who use the AI technology at their workplace must report it to their respective supervisor and uninstall it if necessary, writes the financial news agency, citing people who have seen such specific instructions. In some cases, employees have been completely forbidden from installing it, and this may even apply to family members of military personnel.
The government's reaction clearly shows the hype surrounding OpenClaw in the People's Republic recently. The CNCERT/CC points out that various internet providers have therefore already included hosting offers for OpenClaw in their program. Elsewhere, such services are not yet available; here, one must still install OpenClaw oneself – with all the associated risks. According to the trade magazine The Information, enthusiasm for OpenClaw in China is likely even greater than in Europe or the USA, where the technology caused a stir at the beginning of the year. The government now fears that this entails enormous security risks.
Hobby project triggers hype
OpenClaw originated from a hobby project by Austrian Peter Steinberger; he had it largely programmed by an AI. It is a powerful AI agent that is installed on one's own hardware and gets full access there. It can be remotely controlled via a messenger like Telegram and perform any tasks automatically and autonomously. For example, it can install or delete software and execute it. The technology was originally called "Clawdbot", was renamed "Moltbot" due to the name similarity to Anthropic's AI Claude, and is now operating under the name OpenClaw. In mid-February, Steinberger joined OpenAI. There, he is to develop the next generation of personal AI agents.
Videos by heise
In parallel, OpenClaw has caused a real storm in China, explains The Information. Especially during the holidays around the Chinese New Year, developers worked on building new products from the AI technology. Despite the enormous risks associated with giving an AI full access to a computer, Chinese corporations like Tencent and JD.com have released apps based on it, writes Bloomberg. Local governments have announced subsidies for companies developing products based on OpenClaw. Beijing now apparently wants to get this under control and at the same time prevent a loss of control. However, a ban has been avoided. Nevertheless, tech stocks in the People's Republic have fallen, Bloomberg also writes.
(mho)
