The AI software OpenClaw (commonly known as “Lobster”), developed by Open Source AI, has gained rapid popularity on Chinese social media platforms over the past few months, with “raising lobsters” becoming one of the hottest topics in the AI community. However, there have been reports that the Chinese authorities have issued risk warnings, cautioning state-owned enterprises and government departments against installing OpenClaw in offices for security reasons.
Observers have noted that OpenClaw, often referred to as a more user-friendly version of Siri (AI virtual voice assistant), consumes a large amount of computing power and comes with hidden costs. Despite the encouragement to promote it, this reflects the Chinese government’s strategic layout in AI. Due to security concerns arising from the use of Lobster, the Chinese authorities hurriedly imposed restrictions, leading to contradictions in their policies and regulations.
Originally launched in November 2025, OpenClaw differs from traditional AI chatbots in that, once authorized by users, it can automatically perform tasks such as reading documents, searching for information, coding, and even sending emails, completing complex tasks from start to finish on your behalf. Many engineers see OpenClaw as a “digital worker” capable of practical work.
Recognized by its icon of a red lobster, OpenClaw has been dubbed as the “Lobster.” Currently, several Chinese cloud service providers such as Tencent Cloud, Alibaba Cloud, Baidu Smart Cloud, and Volcano Engine have deployed OpenClaw deployment features.
However, Xue Zongzhi, the procurement manager at TSMC and currently the CEO of a tech company, explained that the rapidly popularized open-source AI Agent platform “OpenClaw” driven by the community and independent developers is essentially your personalized Siri (AI virtual voice assistant), as Siri is considered rather weak. Xue described OpenClaw (Lobster) as an algorithm that customizes and learns to address user queries more interactively.
Regarding the phenomenon of “raising lobsters” in China over the past few months, it was a result of major domestic internet companies, smartphone manufacturers, and local governments eagerly joining the trend, certifying and endorsing the wave of enthusiasm to drive its continuous spread.
In the unstoppable lobster-raising trend, Tencent Holdings’ stock price surged by 7.3% on the 10th. Startups such as MiniMax, identified as the primary beneficiaries of OpenClaw’s proliferation, witnessed their stock prices skyrocket by over 20%, becoming “golden stocks.”
To seize the opportunities in the intelligent economy, local governments have followed suit by introducing policies such as the “Lobster Ten Commandments” and “12 Rules for Raising Lobsters,” offering subsidies for free deployment, computing power, and subsidies for large model deployment fees up to 5 million yuan. The topic of civil servants raising lobsters quickly trended on Weibo.
In response to the encouragement for raising lobsters by local governments, Xue Zongzhi expressed that it may not yield significant returns, as maintaining a 24/7 server incurs substantial data costs that differ vastly from the data volume required by cloud services for millions of users collectively.
Zhao Junshuo, a speaker at the US-China Taiwan Strategy Room, highlighted that commanding the Lobster to perform tasks consumes considerable computational power, similar to engaging AI software to tackle complex problems, which incurs charges accordingly.
He emphasized that while it may appear cost-effective, using such AI software is resource and cost-intensive, with subsidies being offered by the Chinese authorities to consumers using the software. Despite seeming affordable, widespread adoption could lead smaller companies to struggle.
Journalists tracking user bills and interviewing operators have revealed hidden cost traps and security risks lurking behind the AI craze.
With reports of exorbitant daily costs exceeding 2,000 yuan for lobster raisers, some likened the situation to a “digital assassin bill” in the digital era, pointing to escalating expenses and unexpected inefficiencies.
Programmers in Beijing have warned about the emergence of the first batch of OpenClaw victims, cautioning about token usage resulting in unexpectedly high bills, leading to bankruptcy risks.
Furthermore, behind the frenzy of raising lobsters lies significant risks and challenges. Users’ feedback has highlighted issues such as content deletion, privacy breaches, and unauthorized expenses during the lobster-raising process. The initial raisers have been seen urgently uninstalling the software, with some paying hundreds of yuan for assistance.
Wang He, a China issues expert and current affairs commentator, pointed out that Lobster and robots are intertwined as part of the Chinese Communist Party’s strategic layout to establish domination in AI, aiming to surpass the US.
Recent statements by the Chinese Communist Party focused on 2025 being a crucial year for breakthroughs in domestic humanoid robot technology, emphasizing the integration of AI, high-end manufacturing, new materials, and other technologies as core components of the 15-year plan.
However, the Ministry of Industry and Information Technology (MIIT) monitoring detected higher security risks in certain instances of OpenClaw’s open-source AI intelligent body when default or improper configurations were applied, posing high security threats such as network attacks and information leakage.
According to Bloomberg News, Chinese government offices and state-owned enterprises, including major state-owned banks, were recently warned against installing OpenClaw on office devices for security reasons. Those who have already installed the software are required to report to superiors for security checks and potential removal of the program.
Insiders revealed that the ban extended to military personnel and their families.
Wang He noted that the Chinese government is cognizant of the information security risks in using Lobster, prompting strict control measures. The recent US capture of Maduro and killing of Soleimani through AI technology and big data has deeply concerned and frightened Xi Jinping, underscoring the advanced capabilities and mature mechanisms of the US in the field of AI.
“If OpenClaw presents opportunities for China to overtake others, it also poses unprecedented threats to the communist regime. Hence, the primary focus is on safeguarding security, leading to seemingly contradictory policies with underlying internal logic,” outlined Wang He.
In the wave of AI artificial intelligence, the Chinese Communist Party believes it can compete with the US in this realm to surpass the US. However, the reality is that China lags far behind in true AI technology compared to the US, with substantial gaps still present.
In October 2024, the Biden administration imposed new restrictions on China’s investments in AI, semiconductors, microelectronics, quantum computing, and other technologies to protect national security, effective from January 2, 2025.
Wang He emphasized that the US has detached itself from investing substantially in China’s high-end AI sector. Personnel exchanges have been severed, widening the technological disparity between China and the US in the AI domain.
Ultimately, the US leads in funding and pioneering research discoveries in AI and artificial intelligence, fostering original scientific research capacities. While China may perceive itself as having made significant progress and achievements, the gap in technology remains substantial.
In conclusion, the article delves into the complexities of the OpenClaw AI software and its implications on the Chinese technological landscape, shedding light on the challenges and opportunities presented by this innovative platform.