A viral AI framework called OpenClaw has been generating massive public interest recently. On March 6th, nearly a thousand developers and AI enthusiasts queued up outside TENCENT's headquarters in Shenzhen, waiting for assistance from Tencent Cloud engineers to complete cloud installations of OpenClaw. Media reports indicated the event utilized Tencent's Lighthouse cloud service, enabling free installation within five minutes. Queuing began around 10 a.m. with an initial group of over eighty enthusiasts, and by 11 a.m., all available reservation slots had been fully allocated.
Currently, the number of developers and processing cores utilized on Tencent's Lighthouse cloud service has repeatedly reached record highs, with the community of OpenClaw "lobster farmers" surpassing 100,000 users.
In related online discussions, many users reported using OpenClaw for tasks like organizing files and analyzing gold market trends. However, other comments expressed confusion, with some users stating they felt unable to keep up with the rapid pace of new developments.
Some users voiced security concerns, questioning whether granting extensive computer access to the AI was advisable and describing the potential risks as alarming upon deeper consideration.
OpenClaw is an open-source AI agent framework, symbolized by a red lobster icon. Its slogan is "The AI that actually does things." Unlike traditional conversational AI, OpenClaw is not merely a chatbot that answers questions but functions as a digital employee capable of executing tasks. It operates through natural language instructions and can perform practical operations such as file management, browser automation, API calls, and multi-step task orchestration in local or private cloud environments, representing a shift from providing suggestions to taking direct action.
The framework has recently experienced explosive growth, surpassing Linux to become the most-starred project on GitHub in just one hundred days, sparking a widespread "lobster farming" trend.
This surge has prompted major Chinese cloud service providers to respond collectively. Tencent Cloud was the first to launch a dedicated OpenClaw application template within China, offering one-click deployment and a visual configuration panel. Competitors including Alibaba Cloud, JD Cloud, Volcano Engine, and Baidu AI Cloud have subsequently rolled out their own cloud deployment services, competing to lower usage barriers and capture early advantage in the AI agent market.
Furthermore, users seeking to "run the lobster" found that Apple's Mac mini was completely sold out online, with retail stores also reporting no stock. Upon inquiry, staff at Apple's official stores confirmed that the Mac mini was out of stock across all physical locations in the city, suggesting a possible link to OpenClaw's sudden popularity.
The OpenClaw phenomenon has also captured significant attention from AI industry leaders. At a Morgan Stanley conference on March 6, NVIDIA CEO Jensen Huang highlighted the "Agentic AI" inflection point, describing OpenClaw as the most significant recent software release. He noted that OpenClaw achieved greater adoption in three weeks than Linux did over thirty years, becoming the most downloaded open-source software in history. Huang stated that OpenClaw demonstrates AI's potential to operate within highly personalized environments and directly address redundant tasks for everyday users and enterprises, adding that the AI application layer, where OpenClaw operates, currently offers the highest output and returns.
Kai-Fu Lee, former Global Vice President at Microsoft and Google, recently commented that future competition will focus on multi-agent systems, which he believes will act as a bridge connecting the digital and physical worlds.
The buzz around OpenClaw has also attracted investor interest. One month ago, Wang Huiwen, a founding partner at Meituan, issued a call for startups, inviting teams working on OpenClaw-related ventures needing funding to contact him.
Notably, OpenClaw's security aspects have drawn scrutiny. Last week, Summer Yue, AI Alignment and Safety Director at Meta's Super Intelligent Lab, shared an experience where instructing OpenClaw to "confirm before operating" failed to prevent it from rapidly deleting an inbox, noting an inability to stop the actions via phone. Previously, 360 Vulnerability Research Institute analyzed that in typical personal deployment scenarios, users lacking security expertise might expose themselves to various cyber threats.
Comments