A viral artificial intelligence trend sweeping China has rapidly morphed from grassroots experimentation into a full-scale corporate scramble, before drawing swift scrutiny from regulators and state institutions.
At the center of the frenzy is OpenClaw, an open-source “agentic” AI system capable of executing tasks autonomously — from simple commands to complex workflows — prompting comparisons to a new computing paradigm.
Within weeks of its breakout popularity, China’s largest technology firms moved aggressively to stake their claims. Tencent rolled out QClaw, embedding the tool directly into its dominant messaging platform, WeChat, allowing users to issue commands via chat and have tasks executed instantly.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab.
ByteDance, through its cloud division Volcano Engine, launched ArkClaw — a browser-based deployment aimed at enterprise and developer use cases — while Alibaba introduced JVS Claw, a mobile-focused version designed to simplify installation and onboarding. Xiaomi joined the race with MiClaw, currently in closed beta, enabling users to control smartphones and connected home devices through natural language commands.
The speed of adoption has been striking even by China’s fast-moving technology standards. The phrase “raising the lobster” quickly went viral across social media, becoming shorthand for deploying autonomous AI agents to handle daily tasks — from scheduling and coding to managing digital workflows.
In some cities, demand spilled into the physical world. Users reportedly queued outside offices of major tech firms seeking assistance with installation, while others paid freelancers on online marketplaces to set up the software.
What began as a developer-driven experiment quickly evolved into a mass-market phenomenon, highlighting the growing appeal of agent-based AI systems that move beyond passive chat interfaces into active task execution.
Startups moved in parallel with tech giants. Companies such as Zhipu AI, Moonshot AI, and MiniMax released models and frameworks built on top of OpenClaw, with investor enthusiasm driving sharp gains in their share prices following the announcements.
Nvidia And Global Players Join The Push
The momentum has not been confined to China. Nvidia unveiled NemoClaw, an enterprise-grade platform built on OpenClaw, during its Nvidia GTC conference in San Jose.
CEO Jensen Huang framed the shift in stark terms: “Every company in the world today needs to have an OpenClaw strategy, an agentic system strategy. This is the new computer.”
Nvidia’s version emphasizes enterprise safeguards, including network guardrails and privacy routing, reflecting growing concerns about how autonomous AI agents interact with sensitive systems.
However, the initial excitement has been tempered by rising fears over cybersecurity and data exposure.
China’s official vulnerability monitoring body, operated under the Ministry of Industry and Information Technology, issued an early warning that improperly configured OpenClaw deployments could expose systems to cyberattacks or data breaches.
Unlike traditional software, agentic systems often require deep system access to execute tasks — increasing the potential damage if compromised or misused. In recent days, some users have begun uninstalling the software, in some cases paying others to remove it, mirroring the earlier rush to install it.
The response from authorities has been swift.
Government agencies and state-owned enterprises have begun restricting or outright banning the use of OpenClaw on official devices, citing security risks and concerns over uncontrolled data flows. China has been proactive in balancing rapid innovation with strict oversight of technologies that could pose systemic risks.
Since it was launched, OpenClaw has quickly evolved from model development to deployment of autonomous systems. Unlike earlier waves of AI adoption centered on chatbots, agentic systems represent a shift toward software that can independently perform actions — effectively acting as digital workers.
However, this transition carries significant economic implications.
For technology companies, the emergence of agent-based AI opens new opportunities to embed their platforms deeper into users’ daily activities and enterprise operations. But it raises new challenges for regulators, around security, accountability, and control.
While China’s rapid response to the OpenClaw trend reflects the strength of its technology ecosystem, where large platforms, startups, and hardware manufacturers can quickly integrate and scale new innovations, it also exposes vulnerabilities.
The fragmented and open nature of agent-based systems makes them harder to regulate and secure, particularly when adoption spreads beyond controlled enterprise environments into mass consumer use.
But the trajectory of OpenClaw — from viral adoption to corporate land grab and regulatory pushback — mirrors a familiar pattern in emerging technologies. Early enthusiasm drives rapid deployment, often ahead of full understanding of risks, followed by a period of correction as security, governance, and practical limitations come into focus.
For now, the “lobster craze” appears to be entering that second phase. But the underlying shift it represents — toward autonomous, agent-driven computing — is unlikely to reverse.