On August 26th, the Zhejiang Provincial Data Group Co., Ltd. was officially unveiled in Hangzhou. With an initial registered capital of 2 billion yuan, the company is wholly owned by Hangzhou Iron and Steel Group. Its business registration was completed just five days prior, on August 21st.
As a second-tier provincial state-owned enterprise, the Zhejiang Provincial Data Group carries significant weight. Its establishment was coordinated by the provincial party committee and government, with direct guidance from the provincial SASAC and the provincial data bureau. Notably, Wang Jian, an academician of the Chinese Academy of Engineering and director of the Zhijiang Laboratory, has been appointed as chief scientist and attended the unveiling ceremony. A statement made during the ceremony highlighted the strategic importance: this represents a major strategic move by the provincial party committee and government to advance the construction of an innovative Zhejiang and build a highland for AI innovation and development. Why has such high expectation been placed on a second-tier provincial state-owned enterprise?
Born for the Era
To understand the significance of this move, one must first recognize the current landscape. In 2026, the National Data Bureau designated the year as the "Year of Data Element Value Release." The national public data resource registration platform has already registered over 300,000 data items. This signals a clear top-down directive. In June of this year, Zhejiang's Governor Liu Jie emphasized during a research visit to the provincial data bureau the need to actively explore pathways for data element resource and value realization, adhere to market orientation, establish robust data circulation mechanisms, and systematically advance the development of high-quality datasets and language corpora.
Concurrently, AI is undergoing a fundamental shift in its underlying logic. Over the past three to five years, computational power has been the primary driver of AI development; however, as a new cycle begins, while models and algorithms remain foundational, data has emerged as the critical core element propelling AI into its next phase. Whoever possesses high-quality data supply capabilities will hold the initiative in the next round of the AI race. At the unveiling site, Academician Wang Jian recalled first hearing about "artificial intelligence" as a graduate student in the 1980s, when experts told the audience that the next decade would be AI's decade. He waited through one decade with nothing happening. "Now it has finally arrived, and it exceeds everyone's imagination," Wang Jian said with some excitement. He stated that the current moment represents a historical intersection of three great eras—the industrial revolution, the information revolution, and a renaissance—a once-in-a-century technological transformation. Wang Jian's enthusiasm is well-founded. As data becomes AI's new fuel, those who can aggregate, refine, and circulate dispersed data will seize the advantage. This task cannot be accomplished by individual enterprises, and relying on spontaneous market formation would be too slow—it requires a provincial-level orchestrator. This is exactly the underlying rationale for the birth of the Zhejiang Provincial Data Group.
A Zhejiang-Style New Model
The principal official from the provincial SASAC stated at the unveiling that the goal is to create a new model for data resource, value, and commercialization with Zhejiang characteristics, enabling data to truly flow and generate value in the AI era. Supporting this is the core framework of the Zhejiang Provincial Data Group: "One platform + Three major infrastructures + Full-chain operational support." The "Data Element Comprehensive Service Platform" integrates the entire process of public data authorization, aggregation, processing, quality control, rights confirmation, circulation, and monetization. The "Three major infrastructures" include computing clusters, blockchain foundations, and trusted data spaces, alongside five mechanisms covering security compliance, intellectual property, and revenue distribution to ensure safe, standardized, and sustainable data element operations. This system forms a complete closed loop: public data authorization — data governance and development — unified computing supply — vertical model training — multi-scenario model invocation — commercial application promotion — revenue flow back. Currently, this loop is accelerating. "The three key AI elements—data, computing power, and scenarios—are all progressing steadily," stated Liu Shulin, chairman of the Data Group. Specifically, regarding data, the focus will be on leveraging public data to drive industry data; regarding computing, efforts are underway to provide public computing support via a 100,000-card-level computing cluster; regarding scenarios, the next step involves establishing a Zhejiang provincial scenario innovation company. For the vast number of small and medium-sized enterprises (SMEs) in Zhejiang, the implementation of this system signifies clear good news: barriers to accessing data and computing power are expected to be significantly lowered. Market-based data element circulation now has a comprehensive orchestrator, providing more market players with opportunities for participation.
Why Hangzhou Iron and Steel Group?
Looking nationally, Zhejiang is not the earliest province to establish such a group. According to incomplete statistics, over twenty provincial-level data groups exist across the country. Shanghai made its move in 2022 with registered capital of 5 billion yuan; Jiangsu followed in 2024 with 3 billion; Hubei acted last year with 5 billion. Just one week before Zhejiang's unveiling, Shandong Data Group was listed. Different provinces have taken different paths. Shandong's group is led by IT giant Inspur Group, Jiangsu adopted a hybrid model involving provincial finance and provincial enterprises, and Hubei pursued a route of provincial enterprise consolidation. Zhejiang's start is relatively late, its scale not the largest—so where lies its advantage? Zhejiang has chosen a distinctive path: relying on a provincial state-owned enterprise that has completed its digital transformation to wholly establish the group. That enterprise is Hangzhou Iron and Steel. At the end of 2015, Hangzhou Steel's Banshan steel base was fully shuttered, ending nearly 60 years of steelmaking history. It then pivoted towards the digital track, establishing Zhejiang Provincial Data Management Co., Ltd. in 2017, becoming the province's government data open unit—five to six years earlier than most provincial-level data groups. Over the subsequent eight years, Hangzhou Steel built cloud computing data centers, planned over 20,000 cabinets capable of hosting approximately 300,000 servers, developed industry cloud clusters, released proprietary computing hardware brands, and constructed a complete industrial ecosystem covering computing center construction, intelligent computing equipment manufacturing, industry cloud services, data value conversion, and data security assurance. Hangzhou Steel has been listed among the Fortune Global 500 for five consecutive years and is the only provincial state-owned enterprise in Zhejiang primarily focused on the digital industry. The logic for choosing Hangzhou Steel is pragmatic: the Data Group's actual controller remains the provincial SASAC, eliminating the need for a new administrative apparatus and shortening decision-making chains; Hangzhou Steel already owns several mature enterprises that can be directly integrated, avoiding duplicate construction. Data element operations are not about building office towers; they require a triple overlay of technical capability, industry understanding, and compliance expertise. Clearly, relying on an enterprise that has been operating in the data sector for eight years is far more reliable than starting from scratch.
Focused on Artificial Intelligence
The foundation is set. The next question is: what differentiates this company from data groups in other provinces? The answer from Liang Lipeng, general manager of the Zhejiang Provincial Data Group, is unequivocal: artificial intelligence. "The Zhejiang Provincial Data Group is essentially an enterprise serving the development of Zhejiang's AI industry," he stated. The contracts signed on the day of the unveiling offer a glimpse. First came six national AI application pilot bases covering petrochemicals, cultural tourism, earth sciences, healthcare, spatiotemporal geographic information, and embodied intelligence. These bases bear responsibility for the "last mile" of AI technology transitioning from the laboratory to industrialization; the Data Group's role is to provide them with high-quality "fuel." Signing with the pilot bases addresses the question of "where data is used." For example, in the medical field, previously, SMEs wanting to train AI models on datasets for lung cancer or liver cancer had to visit hospitals one by one, a process fraught with lengthy procedures and complex approvals. Now, the medical pilot base integrates data from over 1,500 medical institutions across the province. When enterprises raise demands, the Data Group can directly match and extract relevant data from the resource pool. Another example is multi-industry data fusion. Single-industry data, such as natural resources, national land red lines, and aerospace remote sensing, have limited value when used alone; they need to be combined with multidimensional data from transportation, communications, and other sectors to unlock compound value. Now, the Data Group can coordinate unified access, enabling comprehensive utilization of multidimensional data. Next came digital business ecosystem partners and token factories: SUPCON Group, Mobvista (Daily Interactive), Beijing Silicon Flow, DBAPP Security (Hangzhou Anheng Information), Ant Technology, the China Academy of Industrial Internet, and Zhejiang Xiwang Intelligent Technology, among others, filling roles in computing scheduling, data security, and industrial internet. Signing with these partners addresses "how to use data well." Take data security as an example: previously, SMEs lacked the capacity to build their own security protection systems. Now, the Data Group aggregates security monitoring data centrally and collaborates with DBAPP Security, accumulating monitoring data from over 20,000 enterprises to provide low-cost security services that require no self-built infrastructure. Regarding Zhejiang's potential, Academician Wang Jian offered this assessment: "Having traveled to all provinces across the country, I have come to one judgment. Though Zhejiang is a traditionally resource-poor province, it possesses immense potential to become a strong and major province in data resources. The accumulation over the past decade or more in the digital economy appears particularly crucial against this backdrop."
Disclaimer: This article is republished for the purpose of conveying more information and sharing perspectives. It does not imply verification of its authenticity nor constitute any other form of advice. It serves solely as an exchange platform and does not assume responsibility for copyright. If infringement occurs, please contact us for timely modification or deletion.
Comments