Grace Shao says China has vast data reserves but no breakout independent AI data giant yet

AI Proem author Grace Shao said in a post dated Oct. 9 that China has accumulated enormous volumes of data over time, yet it still has not produced a large standalone AI data company comparable to U.S. players such as Surge AI, Scale AI, and Mercor. She pointed to several structural reasons: frontier model development and upstream data markets took shape earlier in the United States, major Chinese tech companies have long kept data inside closed ecosystems, and corporate and traditional industry data remain fragmented, with state-owned enterprises reluctant to hand data to outside vendors. The post cited data from China’s National Data Administration showing that China generated 52.26 zettabytes of data in 2025, while only 22.13% of stored data was classified as structured. It also referenced an interview by DataYuan with Benyuan Zhishu CTO Lin Zhenya, who estimated that China has around 2,000 to 3,000 data-labeling companies, with the top three accounting for 20% to 30% of the market, versus an estimated 70% to 80% for the top three in the United States. Shao said that figure was an industry estimate rather than an independently calculated market share. The post also mentioned financing for UniPat and Humanlaya, HSG’s investment, a 2026 acquisition involving Benyuan Zhishu, and Haitian Ruisheng’s annual revenue growth.