Growing demand for high-quality data is raising the cost of barriers to data sharing across borders. Sharing data across organisations and jurisdictions generates economic and social value – from improving supply chain resilience to accelerating healthcare research. Yet many organisations that hold valuable data often lack the incentives, governance frameworks or legal certainty to share them. Sharing data has long been an economic and social necessity, but it has intensified as progress in artificial intelligence (AI) from scaling computation alone appears to slow, and increasingly depends on access to high-quality, domain-specific data. Centralised commercial platforms have partially closed this gap, but they require participants to trust a single data intermediary with their data, a condition that is not feasible in all settings.
Data spaces have emerged as a response to this type of challenge, and functionally equivalent approaches exist globally. A data space is a data ecosystem characterised by (i) federated or decentralised governance; (ii) open, standards-based interoperability; and (iii) mutual trust and effective control between data providers and data users over how data are used. Policy efforts on data spaces are often associated with European initiatives, where public investment in data spaces exceeded EUR 1.7 billion as of 2025. However, functionally equivalent approaches, sometimes developed independently and under different terminologies, have emerged. Key examples can be found in Japan (Open Data Spaces, ODS), the United States (Trusted Exchange Framework and Common Agreement, TEFCA) and India (Data Empowerment and Protection Architecture, DEPA). This suggests that multi-party data sharing under federated governance addresses a structural need that is not specific to any single jurisdiction or technical architecture.
Data spaces are best suited for a specific class of use cases involving multi-party data sharing. The analysis identifies four conditions that, taken together, distinguish settings where data spaces offer advantages over common alternatives, such as bilateral contracts, centralised platforms, privacy-enhancing technologies (PETs): (i) data sharing among three or more parties; (ii) competitive or jurisdictional relationships that preclude trusting a single data intermediary; (iii) non-excludable collective value that bilateral contracts alone cannot capture; and (iv) ongoing data flows requiring continuous governance. These features position data spaces as key infrastructures for advancing data free flow with trust (DFFT), as demonstrated by operational use cases in mobility, maritime logistics and supply chain management, where the cumulative conditions of competition, regulatory pressure and cross-border collaboration apply.
Despite significant investment, few data spaces have reached financial and operational viability. Case study evidence from industry-led Catena-X, the Finnish Maritime Data Space (MDS), and Japan's Automotive and Battery Traceability Centre (ABtC) show that data spaces work in practice. The International Data Spaces Association (IDSA) Radar, the largest publicly available registry of data space initiatives, lists 235 entries as of February 2026, nearly triple the 87 recorded in 2022. European initiatives dominate the listings, reflecting the historical concentration of data space activity around EU-funded programmes. Only 25 of these initiatives (one in nine) have reached operational status, and none have achieved the scaling stage associated with financial viability. Most depend on public funding, and few have business models that could sustain operations independently. Once operating at scale, data spaces can more systematically support trusted data sharing across jurisdictions. This has already been demonstrated as a proof of concept by the exchange of supply chain data between Catena-X and the Ouranos Ecosystem across the European Union and Japan.
Establishing a data space faces a bootstrapping problem. The shared infrastructure must precede the network effects that justify it. Moreover, each participant’s onboarding investment must be justified by returns that may not materialise until the ecosystem reaches sufficient scale. This chicken-and-egg dynamic, referred to as the “bootstrapping problem”, is a central barrier to data space viability. In cases where it has been overcome, as in the Finnish MDS, quantifiable use-case value was present from the outset. ABtC in Japan followed the same pattern: original equipment manufacturers and suppliers were anchored by a concrete initial use case (carbon footprint data) before the scope expanded across the battery lifecycle. Where the bootstrapping problem has not been overcome, as with Agdatahub (which declared bankruptcy in 2024), infrastructure was built without demonstrated use-case value sufficient to sustain participation once public grant funding expired.
The most limiting factor may be institutional design rather than funding. Continued funding may itself reduce the pressure on data space ecosystems to develop business models that can sustain them after the subsidy ends. Where funding rewards grant compliance over outcomes, rational actors optimise for the former, a misaligned incentive documented in public venture capital. Overlapping initiatives developing redundant building blocks compound this risk in two ways: duplicative research and development (R&D); and delays driven by possible vested interests, another pattern well documented in technology standardisation more broadly.