Lure of the lobster: Why OpenClaw caught China’s imagination

BEIJING - A lobster craze is sweeping across China – not in kitchens, but in the country’s fast-moving world of artificial intelligence.

The “lobster” is OpenClaw, a newly popular AI agent that has inspired queues, tutorials and no small amount of hysteria.

In early March, hundreds of people gathered outside Tencent’s headquarters in Shenzhen, clutching laptops and hard drives, waiting to get help installing the software. Appointment slots ran out in about an hour.

On e-commerce platform Taobao, freelance technicians offered remote installation for fees of up to $100, a virtual echo of the small shops in Sim Lim Square that used to assemble custom-built PCs during the early days of the internet boom.

Some of the same technicians soon found themselves doing repeat business, hired again to uninstall the program once the initial excitement wore off and users realised it was harder – or riskier – to use than it had first seemed.

The appeal is easy to understand. OpenClaw promises to do more than just talk. Unlike earlier chatbots that answer questions, it can carry out tasks on a user’s behalf – reading files, sending messages, writing code or organising schedules across different applications.

As it can work round the clock and perform a wide range of tasks autonomously, OpenClaw has been touted as an uncomplaining all-in-one sidekick – cheaper, more capable and more dependable than a human assistant.

Created by an Austrian developer and released only months ago, the program spread quickly through programmer circles before spilling into the mainstream.

Few Chinese users seemed overly concerned that OpenClaw was developed overseas. When some warned online that foreign software might contain backdoors, others replied with dry humour that the risk did not feel unusual – and that Chinese users were already accustomed to living with such possibilities in domestic apps.

Cult of technology

At offline meet-ups organised to boost adoption, the mood was febrile and even faintly cult-like. Photos online showed participants wearing hats shaped like red lobster pincers, facing a screen displaying a slogan that declared only those who mastered OpenClaw would earn a “ticket to the Web 4.0 era”.

Some commentators compared the spectacle to scenes in the early 1990s, when crowds gathered on a mountain in Beijing wearing aluminium pots on their heads, believing the metal could help them receive cosmic signals.

That was the era of the qigong craze, when a society keenly aware that it lagged behind its more advanced Asian neighbours developed an intense thirst for science – but not always the ability to distinguish between genuine discovery and wishful thinking.

Comparing the geeks embracing the future of AI to the elderly who sought divination through aluminium pots is unnecessarily harsh. But it captures something real about the emotional force behind today’s enthusiasm.

That urgency has deeper roots in China’s growth trajectory. Over the past four decades, the nation’s rise from poverty to relative prosperity has been driven in part by waves of new technology that created sudden opportunities for those quick enough to act.

Those who leapt into e-commerce by opening stores on Taobao in the 2000s, for example, saw their businesses flourish while many bricks-and-mortar shops were forced to close.

About two decades later, when live-streaming took off, online retailers on Douyin who embraced the format thrived, drawing sales away from competitors who did not. By 2025, sales generated through live-streaming on Douyin had surpassed those of some older e-commerce rivals such as JD.com, reinforcing the sense that missing a technological shift is like letting money slip through one’s fingers.

There is also a deeper cultural instinct at work. China’s openness to the latest technology reflects a longstanding belief that science can transform its fortunes.

In the late 1970s, an abstract mathematical puzzle known as Goldbach’s conjecture became a household name and national obsession after mathematician Chen Jingrun was portrayed as a hero of national revival.

Since then, science and technology have been associated with progress and the promise of catching up with the world. That instinct helps explain why each wave of innovation, from mobile payments to short-video apps, tends to spread in China with unusual speed.

The industry incentive

The excitement around OpenClaw is not driven by individuals alone. China’s technology industry has strong reasons of its own to encourage the rush.

Over the past few years, the country’s biggest tech firms have poured money into AI servers and cloud infrastructure, and now face pressure to generate enough usage to justify those investments.

Unlike chatbots, which people use occasionally, AI agents can run for hours and consume far more computing power. The more widely such tools are deployed, the more demand flows to cloud providers and model developers. That gives the industry a clear incentive to encourage adoption.

The push has also been reinforced by official signals encouraging faster adoption of artificial intelligence, with the policy agenda in 2026 repeatedly calling for “AI-plus” applications across the economy.

China’s leading tech firms – including Tencent, Alibaba, ByteDance and Xiaomi – have wasted little time rolling out OpenClaw-related products, eager to ensure the latest AI trend runs through their own platforms rather than someone else’s.

Cautious authorities

The frenzy has in turn drawn the attention of regulators, who are moving to set limits on how far the technology can spread.

The authorities have warned that OpenClaw’s broad access to files, accounts and networks creates risks ranging from data leaks to cyber intrusions if used without proper safeguards.

Some government agencies and state-owned firms have been told not to install it on office systems, and companies offering related services have been reminded that commercial applications must comply with rules on generative AI, data security and cross-border information flows.

The concern is not only about hacking or privacy, but about the dangers of a powerful new tool spreading faster than it can be supervised in a system that places a high premium on control of information.

Much of the risk comes from the access the software needs. To act independently, the agent must be given broad permissions and allowed to run in the background, making it harder to control. If not carefully configured, it may take unintended actions, from deleting files to trying to bypass China’s internet firewall to trade cryptocurrency, potentially breaking the law.

Even with ready-made plug-ins, OpenClaw is not quite the effortless assistant many imagined. Users still need to tweak settings, define boundaries and watch what the agent is doing – a process enthusiasts jokingly call “raising the lobster”, as if the AI needs to be fed, trained and occasionally restrained.

Still, the OpenClaw mania is real and episodes like this reveal how new technologies spread in China: not gradually, but in surges, driven by policy signals, commercial incentives and a public eager not to be left behind.

That anxiety is easy to understand. In a country where each wave of innovation has reshaped fortunes, missing the next shift can feel like missing a chance at progress itself. The instinct to move quickly – to jump on the bandwagon first and figure things out later — is part of what has powered China’s rise over the past four decades.

Yet, the same instinct also makes it easier to be swept up in hype, especially when companies with billions invested in new technology have every reason to encourage the rush. New tools are often presented as indispensable long before they are fully mature, and before most users have a clear idea of how – or whether – they truly need them.

Sometimes, though, it pays to let the bullets fly for a while – to borrow a line from the iconic Chinese film of the same name. Not every wave in the technology cycle needs to be ridden at once. It is possible to enjoy the excitement of a new technological frontier without surrendering to the fear of being left behind.

When adoption moves faster than understanding, the greater risk may not be falling behind, but being carried along before the technology – and its limits – become clear.