GoITSystems by Dr. M. Gottlieb

Digital Sovereignty and the Architecture of Decentralized Data Economies: Challenges for Information Systems

Digital Sovereignty (AI genereated)
Digital Sovereignty (AI genereated)

Organizations want to share data without unnecessarily limiting their autonomy in their relationships with technology providers and partners. Data spaces address this tension by combining cross-organizational data sharing with common rules, standards, and trust mechanisms. For information systems research, this raises a sociotechnical design question: How can data infrastructures enable collaboration while preserving transparent decision-making and oversight? Research on federated data ecosystems and their governance provides a starting point ().

1. The Challenge: Sharing Data While Preserving Autonomy

Digital sovereignty is not a clearly defined technical state. The term encompasses different political and organizational conceptions of who can shape digital systems, use data, and establish binding rules. Academic research must therefore begin by specifying whose sovereignty is being examined: that of a state, an organization, or an individual. Equally important is identifying the dependencies in relation to which these actors should retain the ability to act ().

This article adopts a working definition of digital sovereignty as an organization’s ability to make informed decisions about its digital resources and put those decisions into practice. The focus is on data access, terms of use, responsibilities, and the ability to switch providers or arrangements. Under this definition, sovereignty does not mean complete independence; rather, it means managing dependencies deliberately and in ways that can be assessed.

For businesses and public-sector organizations, this creates a concrete design challenge: making data available for jointly delivered services while maintaining clarity about responsibilities, permitted uses, and oversight. Data spaces offer an approach that combines technical connectivity with organizational agreements ().

2. What Data Spaces Offer—and Where Gaia-X Fits In

A data space is more than a platform for transferring files. It connects participating organizations through shared rules, technical standards, and data-sharing services. Its defining feature is that data from different providers can be discovered, accessed, and used under agreed-upon conditions. Building these environments therefore requires both technical infrastructure and a viable operating and governance model (; ).

Gaia-X should not be equated with a single data space or a centralized cloud offering. The initiative addresses the frameworks needed for trustworthy, interoperable data and service ecosystems. Individual data spaces also require their own use cases, participating organizations, operating models, and data-use agreements ().

Academic analysis should distinguish among three levels:

  • Technical level: How are data described, transferred, and processed? Which interfaces and identity mechanisms connect participants?
  • Organizational level: Who makes decisions about participation, access conditions, changes, and conflict resolution?
  • Economic level: How are benefits, costs, and risks distributed so that sustained collaboration remains attractive?

This distinction helps prevent technical connectivity from being mistaken for effective collaboration. A data space may be technically feasible yet still fail because decision rights are unclear or its operating model is not viable. The relationship among architecture, governance, and economic value is therefore central to data space design ().

3. Governance at the Core of the Design Challenge

This article focuses on the governance of data infrastructures: the rules, roles, and decision-making procedures through which a data space is organized. These include onboarding new participants, establishing acceptable terms of use, verifying compliance with agreed-upon requirements, and defining conflict resolution procedures. Technical capabilities can support these rules, but they cannot replace agreement on which rules should apply ().

For research purposes, these considerations suggest an analytical framework with four design dimensions:

  • Decision rights: Who is authorized to provide data, approve access, and modify rules?
  • Trust and verifiability: Which attributes of participants and services need to be verifiable?
  • Interoperability: How are technical interfaces aligned with a shared understanding of the data?
  • Viability: Which incentives and resources sustain participation and operations beyond the initial implementation phase?

This framework is intended as a proposed analytical structure, not an already validated measurement model. An empirical study would need to refine these dimensions, connect them to the relevant literature, and operationalize them as observable attributes or measures.

Digital Credentials: Supporting Trust While Recognizing Limitations

One technical building block is the use of digitally verifiable attestations known as Verifiable Credentials. The data model published by the World Wide Web Consortium describes how an issuer’s claims about a subject can be represented in a verifiable form. Such credentials can, for example, support the verification of attributes of participating organizations ().

Three questions must be distinguished: Has the credential remained unaltered? Is its issuer trusted? And does the verified attribute authorize a particular type of access? A verifiable credential does not automatically answer all three questions. Nor does it, on its own, ensure that data will be used exclusively as agreed once they have been made available. Identity verification, access decisions, and control over data use must therefore be treated as distinct design challenges ().

4. Research Design: From Problem Definition to an Evaluable Contribution

The choice of research method should follow from the intended contribution to knowledge. Researchers seeking to develop a solution to a specific design problem may adopt Design Science Research. This approach combines artifact development with a justified evaluation of the artifact’s utility and quality. An artifact may take the form of a model, a method, or a prototype implementation ().

Potential artifacts for data spaces include:

  • A governance model that specifies roles, decision rights, and escalation procedures.
  • A method for selecting appropriate rules for data access and use.
  • A reference architecture that integrates identity verification, access control, and supporting evidence.
  • A demonstrator for testing selected rules within a clearly defined application scenario.

The key is to move beyond simply describing the artifact. A research study must clearly explain the problem it addresses, the requirements informing the design, and the criteria used to evaluate the solution. Implementing a single data space does not, by itself, constitute a generalizable research contribution ().

An Illustrative Research Scenario

One suitable scenario could involve several public-sector organizations contributing selected data to a joint planning effort. Rather than examining whether data spaces are generally superior to other arrangements, the study would address a clearly bounded design question: How should onboarding procedures, access decisions, and responsibilities be structured so that participants can share data under transparent conditions?

A proposed research design could begin with interviews and document analysis to identify requirements and conflicting interests. These findings could then inform the development of a governance model to be tested in a pilot project. Evaluation criteria would need to be established in advance and could include:

  • The effort and time required to onboard new participants.
  • The clarity and traceability of access decisions.
  • The effort required to modify terms of use.
  • The handling of rule violations and conflicts.
  • Participating organizations’ perceived benefits and perceived autonomy.

Technical tests and organizational assessments should be analyzed separately before their findings are examined in relation to one another. This makes it possible to distinguish difficulties arising from implementation issues, unclear rules, or conflicting interests. The criteria listed above are suggestions for an evaluation design; their suitability must be justified for the specific use case.

5. Research Questions Rather Than Blanket Claims About Research Gaps

Research on data spaces already encompasses technical concepts, governance approaches, implementations, and use cases. A blanket claim that robust empirical studies are lacking would therefore require a systematic review of the literature to be adequately substantiated ().

For theses and doctoral dissertations, precise questions are more useful than broad but unsupported claims about research gaps. The design challenges discussed above suggest several possible research questions:

  • How does the allocation of decision rights influence organizations’ willingness to provide data?
  • Which governance rules reduce the coordination effort required to onboard new participants?
  • How do costs, benefits, and perceived autonomy change over the course of a data space project?
  • Under what conditions do standardized credentials support trust building among organizations?
  • How can AI-enabled analytics be integrated into a data space in ways that allow adherence to agreed-upon access and usage conditions to be verified?

Whether these questions reveal a defensible research gap must be established for the specific question selected. This requires transparent documentation of the search strategy, databases, search period, and inclusion and exclusion criteria. Only then can researchers substantiate which relationships, application contexts, or developments over time remain insufficiently studied.

6. Implications for Research and Practice

For organizations, the central design question is not simply whether to adopt a centralized or decentralized architecture. Rather, it is which combination of infrastructure, rules, and responsibilities best fits the use case. The data space literature highlights the need to consider technical and economic design together; it does not establish the general superiority of any one architecture ().

For students and doctoral researchers, the scholarly contribution lies in clearly connecting the problem, research question, theoretical framework, and evaluation. A focused investigation of decision rights, trust mechanisms, or conditions for participation is more informative than an attempt to measure “digital sovereignty” as a whole.

Viewed this way, digital sovereignty becomes a design challenge that can be evaluated: Which decisions can participants actually make? Which rules are effectively implemented? And which dependencies remain? Data spaces provide a relevant setting for investigating these questions—not as ends in themselves, but as a form of cross-organizational collaboration that can be deliberately designed.

References

Publications

Social Networks

Newsletter - Anmeldung