- Hangzhou is building state-backed infrastructure to unlock the value of public data for AI and business
- The challenge is balancing tighter data controls with open, competitive access
Hangzhou launched its municipal data group with 5 billion yuan ($744 million) in registered capital in March 2026. Four months later, Zhejiang Province followed provincial-level data platform with 2 billion yuan in registered capital.
The timing is unlikely to be coincidental. As data becomes an increasingly valuable input for AI, China faces a basic question: how can data be put to work without breaching the country’s security and compliance rules?
A red line around data flows
China’s Data Security Law, which took effect in 2021, established a framework for protecting data and assigning security responsibilities across government and key industries. It allows data to circulate, but within a defined compliance framework.
The challenge is greatest for public data. Governments hold vast datasets with potential value for AI and businesses, but much of that information also touches economic security, public interests or other sensitive areas.
The law sets the security principles, but leaves considerable room for local governments to determine how public data can actually be commercialized.
That creates a gap between data value and data access.
Building the compliance channel
Hangzhou’s municipal and the provincial data groups are designed in part to fill that gap, acting as an intermediary between government-held data and the commercial market.
Their role is to enable data to be classified, anonymized, secured and reviewed before being turned into products or services.
The Hangzhou Data Exchange, which is wholly owned by the Hangzhou Data Group, has already experimented with data trusts, data-asset accounting and data-backed financing.
The broader objective is to make public data usable without simply handing it over.
The new digital infrastructure
This gives data groups a significance beyond their corporate balance sheets.
The most valuable asset they control may not be the data itself, but the gateway through which public data enters the economy.
That increasingly resembles traditional infrastructure. Roads, power grids and water systems underpinned previous generations of urban development; computing and data networks underpin the digital economy.
In the AI era, a third layer may be emerging: the compliance network.
Computing determines how much AI can process; data determines what it can learn; compliance determines what data can legally be used.
The risk on the other side
But concentration creates its own risk.
If government-backed platforms become the main gateways for public data, they could also accumulate significant power over its access, pricing and allocation.
That could create a new digital barrier, particularly for smaller companies that cannot afford the costs of data access and compliance.
The real test for Hangzhou’s model, therefore, is not simply whether it makes data flows safer. It is whether it can make data secure and accessible at the same time.
China has drawn the red line. The next challenge is ensuring that the infrastructure built inside it does not become a new monopoly.
Header image credit: Growtika/Unsplash
