字号·· | 护眼
南华早报

中国的人工智能竞赛需要数据燃料。一家新的国有巨头能否提供?China’s AI race needs data fuel. Can a new state-owned giant deliver?

点「原文对照」整页切到原文,或双击某段只看那段的原文。

北京正在进行一项雄心勃勃的尝试,旨在通过成立一个中央管理的国家数据集团,将中国庞大的数据海洋转化为人工智能和经济增长的燃料,尽管专家警告,其成败将取决于是否能打破根深蒂固的数据共享壁垒。

Beijing is making an ambitious bid to turn China’s vast ocean of data into fuel for artificial intelligence and economic growth with the launch of a centralised national data group, though experts warn its success will depend on tearing down entrenched data-sharing barriers.

该企业于10月20日成立,为中央管理的国有企业,总部设在河北省雄安新区,这标志着北京意在将其数字资源转化为竞争优势。

The venture – established on October 20 as a centrally managed state-owned enterprise headquartered in Xiongan New Area, Hebei province – signals Beijing’s intent to translate its digital trove into competitive advantage.

据《南华早报》上周报道,副总理丁薛祥在揭牌仪式上将此举称为旨在增强中国在人工智能和数字经济领域竞争力的“重大战略决策”。

Speaking at the inauguration ceremony, Vice-Premier Ding Xuexiang described the move as a “major strategic decision” aimed at strengthening China’s competitiveness in AI and the digital economy, the South China Morning Post reported last week.

清华大学现代国企研究院研究主任周丽莎表示,通过在国家层面统筹数据资源,该集团有望降低数据共享和使用成本,加速其转化为具有商业价值的资产,并支持中国经济发展。

By coordinating data resources at the national level, the group was expected to lower the costs of sharing and using data, accelerate its transformation into commercially valuable assets, and support the development of China’s economy, said Zhou Lisha, research director at Tsinghua University’s Institute for Modern State-Owned Enterprises.

周丽莎在清华大学修忠学院近期发布的评论中表示,此举有望通过释放中国庞大数据资源的价值来推动经济增长。

The move was expected to help drive economic growth by unlocking the value of China’s vast data resources, Zhou said in comments recently published by the university’s Xiuzhong College.

这一举措凸显了北京加强支撑其不断壮大的人工智能产业基础设施建设的决心。

The initiative underscores Beijing’s intensifying drive to ramp up the infrastructure underpinning its growing AI industry.

中国科学院大学教授孙毅表示,新集团有助于建立统一的数据标准,统筹数据采集和采购,为国产人工智能大模型和具身智能应用(包括机器人技术)提供高质量数据集。

Sun Yi, a professor at the University of Chinese Academy of Sciences, said the new group could help establish unified data standards and coordinate data collection and procurement, providing high-quality data sets for domestic AI models and embodied AI applications, including robotics.

据官媒《人民日报》周一报道,孙毅表示,这些努力有望降低训练人工智能模型的总体成本。

Such efforts could lower the overall cost of training AI models, Sun was quoted as saying in a report by state media People’s Daily on Monday.

北京正以一批大型国有企业为后盾支持该实体,其中包括该国主要的电信运营商和电网运营商。国海海通证券分析师在上周的一份客户研报中表示,这种架构可能使其能够利用其支持者现有的资源和基础设施。

Beijing is backing the entity with a slew of mega state-backed enterprises, including the country’s major telecoms carriers and power grid operators. That structure could allow it to tap into its backers’ incumbent resources and infrastructure, analysts at Guohai Haitong Securities said in a clients’ note last week.

2026年8月25日,中国国务院副总理丁薛祥在位于中国东部安徽省合肥市中国科学技术大学了解人工智能在科学研究中的应用。新华社发Chinese Vice Premier Ding Xuexiang learns about the application of AI in scientific research at the University of Science and Technology in Hefei, east China’s Anhui province, on August 25, 2026. Photo: Xinhua

中国南方电网数字化部负责人、该数据集团股东之一的n表示,公司将把其能源领域数据共享基础设施与该集团的国家级平台相连接,以促进数据流通并开发融合算力与电力资源的应用,据《人民日报》周一报道。

n, head of the digitisation department at China Southern Power Grid, one of the data group’s shareholders, said the company would connect its energy-sector data-sharing infrastructure with the group’s national platform to facilitate data circulation and develop applications integrating computing power and electricity resources, according to Monday’s report by People’s Daily.

这一战略部署正值快速扩张之际。国家数据局编制的统计数据显示,2025年中国数据产业同比增长15.7%,达到6.78万亿元人民币(约1.02万亿美元),从事该领域的企业数量达48.2万家。

The strategic push comes amid rapid expansion. China’s data industry grew 15.7 per cent year on year to 6.78 trillion yuan (US$1.02 trillion) in 2025, with the number of companies operating in the sector reaching 482,000, statistics compiled by the National Data Administration showed.

然而,一些专家已发出潜在逆风的警示,指出该集团必须应对碎片化的数据标准、不明确的所有权,以及协调不同企业和政府机构所持有资源的困难。

Yet some experts have flagged potential headwinds, warning that the group must navigate fragmented data standards, unclear ownership rights and the difficulty of coordinating resources held by different companies and government agencies.

北京智库盘古智库高级研究员于凤辉表示,数据共享和商业化的长期障碍依然存在,包括交易保障措施不足以及区域交易平台标准不一。

Yu Fenghui, a senior researcher affiliated with Pangoal Institution, a Beijing-based think tank, said long-standing barriers to data sharing and commercialisation remained, including inadequate safeguards for transactions and inconsistent standards across regional trading platforms.

于凤辉在《证券时报》近期的一份报道中被引述称,行业的主要挑战已从缺乏法规转变为碎片化的规则阻碍了数据跨区域、跨行业流动。

The industry’s main challenge had shifted from a lack of regulations to fragmented rules that hindered data flows across regions and industries, Yu was quoted as saying in a recent report by the Securities Times.

除了技术和监管障碍,新数据集团还面临着一场体制层面的平衡术。

Beyond technical and regulatory hurdles, the new data group faces an institutional balancing act.

官方背景的《人民邮电报》上周发表的一篇社论警告称,该集团面临的最大挑战将是在其国资背景的角色与市场化运营之间取得平衡。

An editorial published by the state-backed People’s Posts and Telecommunications News last week warned that the group’s biggest challenge would be striking a balance between its state-backed role and market-oriented operations.

社论称:“过度的行政干预可能会挤压民营企业的创新空间,而过度追求商业利润则会损害其初心使命。”

“Excessive administrative intervention could squeeze out innovation by private enterprises, while an excessive pursuit of commercial profits would undermine its original mission,” the editorial said.

社论还指出,该集团应通过专注于基础设施、数据共享标准和重大应用,充当现有企业的“赋能者,而非替代者”。

The group should act as an “enabler rather than a replacement” for existing businesses by focusing on infrastructure, data-sharing standards and major applications, it added.

阅读媒体原文 ↗