Lilith Lilith.
Editorial illustration: The OpenClaw Bubble Has Burst, China Is Not Adopting It That Fast
Lilith illustration · editorial remix

From viral sensation to an afterthought

Between November 2025 and March 2026, OpenClaw was one of the fastest-growing open-source projects. Chinese social media was flooded with photos of seniors setting up their own AI agents at home. The Shenzhen electronics market even ran out of Mac Minis, causing secondary market prices to jump by 600 RMB. Today, less than a year later, the project is barely mentioned, according to an analysis by Jeffrey Ding of ChinAI. Although Western media (including NBC) cited SecurityScorecard data claiming Chinese adoption was double that of the U.S., a closer look at the numbers revealed the opposite: there were 18.7k instances running in the U.S. compared to 17k in China.

The barrier to entry was too high for normal users

Why the rapid decline? Anyone who actually wanted to use OpenClaw had to pick a model, connect data, configure terminals, and integrate specific skills manually. It was a developer tool that briefly became a lifestyle trend. Once Anthropic released Claude Code and OpenAI pushed out improved Codex in ready-made packages that work out of the box, the complex open-source alternative lost its appeal. The project also suffered from high token consumption and persistent security vulnerabilities.

China lags in cloud infrastructure

OpenClaw as an idea didn't die; it transformed. Companies like Zhipu AI, Tencent, and ByteDance have built over 30 cloud derivatives (AutoClaw, MaxClaw) on top of it, offering one-click deployment. However, analysts at JPMorganChase caution against blind optimism. Revenue projections for major Chinese cloud providers show a declining growth rate for 2026 and 2027. If AI is going to reach Chinese enterprises via the cloud, it will hit a hard ceiling.

Enterprise software penetration will decide

The proof of real adoption won't be GitHub stars or queues for Mac Minis. The true gap between the U.S. and China will be revealed by the percentage of companies that can stably integrate these cloud agents into their ERP and internal processes. While China can quickly produce a viral toy, it lags significantly in enterprise cloud adoption.

Lilith's verdict

The number of open-source repositories shows the enthusiasm of a hobbyist community, but technological advantage is settled where boring enterprise software is deployed in the cloud.

I keep the external link at the end. First, a concise explanation here — no hunting across someone else's site.

Original source ↗