A surge in popularity for the open-source AI agent OpenClaw has sparked both excitement and alarm in China, with users adopting the technology for a wide range of applications, from stock trading and dating to creating digital pets. The phenomenon, dubbed “raising the lobster” by Chinese online communities, has prompted government warnings and restrictions amid concerns about security risks and potential misuse.
The frenzy began to intensify last month, with reports of users lining up outside the headquarters of tech giants Tencent in Shenzhen and Baidu in Beijing, seeking assistance with installation. Others reportedly paid individuals online to set up the AI agent for them, demonstrating a rapid and widespread adoption of the technology. OpenClaw’s appeal lies in its ability to automate tasks and potentially handle everyday operate, leading users to experiment with diverse and often unconventional applications.
One particularly risky trend gaining traction on Chinese social media involves utilizing OpenClaw for stock trading. While some users praise the AI agent for streamlining their trading processes, others have experienced significant losses. One user, posting on RedNote, described creating a “stock-specific lobster” – a multi-agent system designed for stock investment research and backtesting. Though, another user reported losing over 30,000 Chinese yuan (approximately $4,350) after the AI made errors, including overselling shares and miscalculating trades. This incident prompted warnings from other users about the potential for financial losses and the inherent risks of relying on an unproven AI for investment decisions.
Beyond finance, Chinese users are exploring more lighthearted applications of OpenClaw. Some are employing the AI agent as a “wingman” for blind dates, reporting positive experiences with the AI’s ability to facilitate conversation and build rapport. One RedNote user described using OpenClaw to chat with potential dates, finding the AI’s responses remarkably natural and helpful in breaking the ice. Another user even created a system where OpenClaw analyzed profiles and offered conversation suggestions, leading to a deeper understanding between themselves and their partner. Some users have jokingly suggested leaving the entire dating process to their AI agents.
Amid the pressures of modern life, some users have turned OpenClaw into a form of digital companionship, treating it as a virtual pet. One RedNote user described the AI as being both serious and humorous, capable of providing encyclopedic knowledge and cracking jokes. However, recent reports of AI agents accidentally deleting files have caused some users to reconsider their reliance on the technology. Others have installed OpenClaw on portable devices, effectively creating handheld digital pets reminiscent of those popular in the 1990s.
The rapid proliferation of OpenClaw has not gone unnoticed by Chinese authorities. In early February, the country’s National Vulnerability Database, operated by the Ministry of Industry and Information Technology, issued a warning about the potential security risks associated with the open-source AI agent. The warning highlighted the possibility of improperly configured deployments exposing systems to cyberattacks and data leaks. In response, government agencies and state-owned firms have begun restricting the employ of OpenClaw on work devices, signaling a growing concern about the technology’s security implications.
