Governments have funded hundreds of data spaces—why are so few operational?
An OECD review finds only 25 of 235 registered initiatives operational and none at its financially viable scaling stage. The report argues that useful shared data needs a concrete use case before infrastructure spending.
The short version
- What happened
- The OECD analysed more than 200 initiatives, case studies of successes and failures, a 235-entry public registry and an expert workshop.
- Who it affects
- Better-designed data sharing could support healthcare research, supply-chain resilience and public services without forcing every participant to surrender control to one platform.
- What remains unproven
- The report is a policy synthesis and case comparison, not a controlled impact evaluation.
Keep in mind
The report is a policy synthesis and case comparison, not a controlled impact evaluation.
Registry coverage is self-selecting and historically concentrated in European programmes; it is not a census of every comparable system worldwide.
Operational status and financial viability do not by themselves prove social value, safety or equitable access.
Source analysis · 1 linked source. Primary analysed source date: 20 October 2026.
Read the full report · 3 min The evidence, methods, implications and limitations
A large pipeline with very few mature systems
A new OECD report says the infrastructure needed to share high-quality domain data for AI and other services is often funded before its users have a reason to join. The 81-page review draws on more than 200 initiatives, success and failure case studies and an expert workshop. It also analyses the International Data Spaces Association's public registry, which held 235 entries in February 2026—nearly three times the 87 recorded in 2022.
Only 25 entries, about one in nine, were operational, and none had reached the report's scaling stage associated with financial viability. Public investment in European data spaces exceeded €1.7 billion by 2025, but the OECD says many initiatives still depend on grants and lack a model that can sustain participation after support ends.[1]
Start with a shared problem, not a platform
The report says a data space is most defensible when at least three parties need continuing data flows, cannot accept one trusted intermediary, create collective value and require ongoing governance. It points to operational cases in European and Japanese supply chains and Finnish maritime logistics. By contrast, Agdatahub entered bankruptcy in 2024 after grant funding ended without sufficient demonstrated use-case value.
The OECD's eight policy considerations include testing structural need and practical viability before investment, separating common technical foundations from ecosystem funding, supporting the cost of participant onboarding and favouring cost-sharing or membership models over speculative data monetisation. These are policy conclusions from a comparative synthesis, not causal estimates. Stronger evidence would track survival, participation, costs, cross-border exchanges and measured public benefits over time.[1]
The distinction that matters
The report's viability test
A data space is not the default answer to every data-sharing problem.
Need
Confirm three or more parties and no trusted intermediary
Establish a structural case
Value
Name the shared use case and anchor participants
Give organisations a reason to join
Sustain
Fund onboarding, governance and a workable model
Move beyond grant-backed infrastructure