This post is part of Global Voices’ April 2026 Spotlight series, “Human perspectives on AI.” This series will offer insight into how AI is being used in global majority countries, how its use and implementation are affecting individual communities, what this AI experiment might mean for future generations, and more.

China’s latest five-year plan, released in early March 2026, reasserts its goal of leading in artificial intelligence (AI) development by 2030 through its “AI+” initiative, which fosters the integration of AI across all major economic sectors. While it is too soon to predict the outcome of the AI race, AI adoption in China is increasingly visible beyond the tech sector as usage among regular citizens has exploded with the emergence of OpenClaw.

Released in November 2025, OpenClaw is an open-source and locally run AI agent developed by Austrian coder Peter Steinberger. Once set up on a computer, the virtual assistant can connect to LLMs like ChatGPT and Claude and perform tasks such as replying to emails, managing calendars, booking hotels, and managing social media accounts. Users can also install “skills” from platforms such as ClawHub to add new functions, such as controlling electronic appliances.

After the Austrian programmer joined OpenAI and moved his OpenClaw project to an open-source foundation in mid February 2026, major Chinese cloud companies rushed to create their own version of “claw.”

Around February 2025, Chinese social media exploded with promises that OpenClaw would revolutionize users’ daily lives — claiming that regular citizens could now create their own AI agents to handle their work and grow their businesses. Inspired by OpenClaw’s lobster logo, users coined the phrase “raising a lobster” (養龍蝦) to promote the adoption of the AI agent. The phrase quickly went viral and helped spur its adoption. However, in less than a month, many have woken up from the “lobster fever” in the face of the harsh reality that creating a local AI agent is often too time-consuming, difficult, and expensive for non-technical people.

Claw adoption in China

In March, Tencent launched a Clawbot on WeChat that allows users to manage email, calendar, and files via natural language in chatrooms. It also launched “Work Buddy,” a desktop version of OpenClaw, and “QClaw,” a one-click installation of “claw” on mobile devices. Alibaba, a massive Chinese e-commerce platform, launched “Wukong,” a platform that connects to various AI agents to handle complex business tasks such as research, editing, and documentation. Baidu, JD.com, and others also introduced their own versions of “lobster.”

To promote the new AI features, these tech giants hosted physical meetups in early March to help students and their corporate fans set up OpenClaw on their desktops. News sites and social media platforms were flooded with footage of crowds outside Tencent and Baidu’s headquarters, hyping up “lobster fever.”

Moreover, many cities, including tech hubs in Shenzhen and Wuxi, have added fuel to the craze by announcing subsidies of CNY 2 to 10 million (approximately USD 290,000 to 1.4 million) to attract startups and developers working on open-source projects.

Across Chinese social media platforms, OpenClaw fans presented the “lobster” as an efficient assistant, capable of performing complex tasks such as video editing, responding to customer inquiries, reviewing and grading student assignments, and trading stocks around the clock with simple, natural-language instructions. Some influencers even claimed that AI agents will eventually replace computer applications and urged their followers to jump on the trend before it is too late.

Lobster fever has driven many to upgrade their hardware, to the point that computers equipped with advanced chips, such as the Mac Mini, were temporarily out of stock in some cities. On shopping platforms, tens of thousands of OpenClaw textbooks have been sold in a month, and some tech-savvy individuals have made substantial profits by providing online assistance to help newbies install and train their AI agents.

However, as more and more people have picked up the tool, they have faced the harsh reality about the cost of “raising a lobster.” A tech-blogger, Yuan Guoxing, spelled out ordinary users’ frustration on Weibo:

你以为装上“龙虾”，等于请了24小时替你打工的员工？实际是你要花大量时间，去调整那个啥也做不了的龙虾。最关键的是，你得拼命打工，才能养得起它。这可能是2026年最讽刺的事。

很多人以为，养“龙虾”是零成本？太天真了。首先，你要给龙虾安个家：用自己电脑，卡到崩溃；买一台Mac Mini至少要4300元，还被抢断货。租个入门云服务器，每月50-200；这还只是房租。

真正的吞金兽，是虾粮——也就是token。AI不吃饭，只吃token。用便宜国产模型，每月30-80，勉强饿不死；想干点正经活，一天几十、上百；敢碰GPT等顶配模型，一天几百元，一个月可能上万。[…] 装“龙虾”，怕养不起；不装，怕掉队。普通人，该怎么办？

Do you think installing ‘Lobster’ is like hiring an employee who works for you 24/7? In reality, you’ll have to spend a ton of time training that useless lobster. Most importantly, you’ll have to work like a dog to raise it. The whole thing could be the biggest irony of 2026.

Many people think running ‘Lobster’ is free. This is too naive. First, you need to set up a home for the lobster: it crashes on an average machine like your PC; you need to spend at least CNY 4,300 (USD 675) on a Mac Mini, which is sold out everywhere. If hosting it on a cloud server, you need to spend CNY 50–200 (USD 7–29) per month — and that’s just the lobster’s housing rent.

The real money-eater is the lobster’s food — that is, tokens. AI doesn’t eat food — it eats tokens. Using a cheap domestic large language model costs CNY 30–80 (USD 4.3–11.6) a month, barely enough to keep it alive; if you want to do some real work, it costs tens or even hundreds of yuan a day; using top-tier models like ChatGPT costs hundreds of yuan a day — possibly tens of thousands a month. […] If you install a lobster, you worry you can’t afford raising it; if you don’t, you’re afraid of being left out [of the market]. What’s an ordinary person supposed to do?

A token is the smallest unit of data that an AI model processes or generates, and users must pay for it. Since March, token consumption by Chinese LLMs has surpassed that of U.S. models thanks to lobster fever. If a user chooses to run an AI agent on a cloud server, they must pay a monthly fee for the service. Raising lobsters means feeding the development of Chinese LLMs and AI platforms.

In addition to personal or even business secrets in chats and emails.Reportedly, some anxious newbies had to pay extra to remove the AI agent from their devices.

Even state-funded media outlets found the “lobster fever” overblown and intervened to cool it down, as the current marketing narrative has fuelled a widespread anxiety that ordinary people will be left out of the job market and the technology-driven economy unless they master AI agents. Chinese authorities also stepped in to warn users about potential security risks, including malicious skill plugins and prompt-injection attacks embedded in malicious websites, which can manipulate AI agents into performing unauthorized tasks. This has driven many users to switch to Chinese lobster versions and platform services.

While most media outlets see the “lobster fever” as further proof of China’s success in creating a profitable model for AI technology, whether such “tokenomics” can sustain and boost domestic consumption remains uncertain. For ordinary people, popular Weibo user Fan Deng’s advice seems more than appropriate:

养龙虾，不如先养好自己。

Instead of raising a lobster, take care of yourself first.