“Digital Employees” Enter Steel Mills: Efficiency Revolution Begins, Safety and Cost Challenges Remain
Recently, the open-source AI agent "OpenClaw" (commonly known as "AI Lobster") has rapidly gained popularity due to its powerful autonomous execution capabilities, with its GitHub stars exceeding 250,000, marking the evolution of artificial intelligence from "conversational tools" to "digital employees." This trend has swept into the industrial sector, with Shagang Group pioneering the deployment of "Steel Lobster" in its heavy plate workshop, achieving fully automated steel transfer in just 18 seconds per operation, demonstrating the enormous potential of AI agents to empower traditional manufacturing.
However, hidden risks lurk beneath the surface of this feast. On the one hand, data security has become a primary challenge: large AI models consume massive amounts of industrial data, and if vulnerabilities are exploited, core processes, trade secrets, and even nationally sensitive information will be at risk of leakage. On the other hand, the cost of token consumption is evolving into a heavy economic burden—the characteristics of AI agents, such as multi-tool invocation and long contextual memory, are causing token consumption to grow exponentially, with IDC predicting a compound annual growth rate of 3418%. Heavy users can incur daily costs ranging from $900 to $3000, and coupled with recent price increases from major model vendors (such as Tencent Cloud's Hunyuan series, which saw an increase of over 460%), enterprise operating costs face significant uncertainty.
In response to the aforementioned risks, the Ministry of Industry and Information Technology recently released a "Six Dos and Six Don'ts" security guideline, emphasizing adherence to the principle of least privilege, strict control over internet exposure, and careful review of third-party "skill packages." Wang Xiaofan, a member of the National Committee of the Chinese People's Political Consultative Conference, called for clarifying application boundaries and ensuring security compliance while encouraging applications; Li Mengjiao pointed out that AI technology providers possess advanced tools, and data permissions must be given the highest level of attention.
For steel companies, it is essential to seize the opportunities presented by intelligent transformation, but also to build a "double insurance": first, establish a strict data hierarchical management and security review mechanism to properly manage the "key" of core data; second, clearly understand the economic benefits of tokens, adopt a core scenario-first strategy, and introduce intelligent scheduling to achieve a dynamic balance between cost and benefit. Only in this way can companies effectively manage this "super steward" and achieve steady and sustainable development in the wave of intelligent transformation.
Tags: OpenClaw, AI intelligent agents, steel industry, data security, token cost, Shagang Group, Ministry of Industry and Information Technology,
