OpenClaw Goes Viral in China's AI Trading Circle: Retail Traders Use AI to Trade Cryptocurrencies, Some Earning Nearly $3000 in 48 Hours

On March 13, it was reported that the open-source AI agent framework OpenClaw has recently become popular in China, attracting attention from the tech industry and investment communities. Within less than four months, this AI agent system has gained over 260,000 GitHub stars. Many Chinese developers and retail investors are beginning to experiment with it for cryptocurrency trading and financial market analysis, sparking a wave of “AI automated trading experiments.”

According to multiple Chinese tech media outlets, OpenClaw’s influence has expanded from the developer community to a broader audience. Engineers from major tech companies like Tencent and Baidu frequently hold installation events, and office workers, students, and even retirees are trying to deploy their own AI agent systems. Meanwhile, concept companies related to AI agents are also attracting capital attention; for example, Minimax Group’s stock price has risen over 550% in two months.

In the official marketplace ClawHub for OpenClaw, there are now over 300 financial and investment skill plugins. Some developers are attempting to use AI to predict markets and trade cryptocurrencies. For instance, a widely circulated case shows an AI trading robot that regularly scans and predicts market data, combined with weather, sports injuries, and on-chain sentiment analysis, turning $50 into $2,980 within 48 hours. Another robot account, “0x8dxd,” is said to have executed over 20,000 trades, with a total profit of about $1.7 million.

However, the reality is not always so ideal. A two-week quantitative experiment recorded on Cnblogs suggests that OpenClaw functions more as an information analysis tool rather than a reliable automated trading system. The hallucination problem of LLM models may trigger erroneous trades during extreme market conditions, and API delays of 1 to 10 seconds pose high risks during market flash crashes.

Security concerns are also prominent. The “ClawHavoc” supply chain attack at the end of 2025 led to over 1,100 malicious skills being embedded in ClawHub. Security firms Koi Security and SlowMist disclosed that some plugins are specifically designed to target crypto wallets and digital assets. Bitdefender Labs further pointed out that about 17% of third-party skills pose a risk of digital asset theft.

The Chinese Ministry of Industry and Information Technology subsequently issued security alerts, requiring organizations to strictly review permission settings and disable unnecessary network access. Several Chinese universities have also restricted the use of OpenClaw on campus networks.

Market sentiment is beginning to shift. In mid-March, “Uninstall OpenClaw” became a trending keyword on Alibaba’s second-hand platform Xianyu, with some users even paying to have the system removed. A seller in Shanghai stated that each uninstallation costs 299 RMB, and multiple orders have been completed.

Although OpenClaw lowers the technical barrier for AI quantitative trading, industry experts generally believe that AI tools can improve information collection and analysis efficiency, but real investment decisions still need to be made by humans. In exploring the integration of AI trading and the crypto market, opportunities and risks continue to coexist.