Digital sovereignty is not a call for Europe to build everything alone or close itself off from global innovation. It is the practical ability of education institutions, educators and learners to understand, choose, adapt, audit and leave the digital systems on which they rely.
One simple test of digital sovereignty
At the recent EU EdTech for Sovereign Education Futures workshop in Brussels, one question offered a particularly useful way of thinking about digital sovereignty: What would it cost an institution — financially, technically and pedagogically — to stop using one of its everyday digital tools tomorrow?
The answer tells us a great deal. When an institution can export its data, replace a component, preserve its learning resources and continue its work with another provider, it retains agency. When leaving would mean losing records, integrations, content, staff knowledge, workflows or access to essential services, the institution is operating inside a system of dependency.
This does not automatically mean that the technology is poor or that the provider has acted irresponsibly. It means that important future choices may no longer be fully controlled by the education institution. For schools, universities and public education systems, this is not only an information technology problem. It affects institutional autonomy, competition, pedagogy, data protection, accessibility and the ability to innovate.
Sovereignty is not isolation
Digital sovereignty is sometimes presented as a choice between European and non-European technology. That is too narrow. Europe will continue to work with international providers, researchers and technology ecosystems. Complete technological self-sufficiency is neither realistic nor necessarily desirable.
The more useful question is: Under what conditions can any provider participate in European education without capturing its data, infrastructure, pedagogy and future choices?
A European company can also create lock-in. A global provider may be able to meet strong European requirements. Nationality matters in relation to jurisdiction, supply-chain exposure and strategic risk, but it should not replace proper evaluation.
For education, sovereignty should therefore be understood as a set of practical capabilities:
- The ability to understand how a system works and how it uses data
- The ability to choose between meaningful alternatives
- The ability to connect systems through open and documented standards
- The ability to adapt technology to educational needs
- The ability to audit high-impact decisions and AI behaviour
- The ability to export information and leave a service
- The ability to maintain pedagogical and institutional autonomy.
These capabilities connect four inseparable dimensions: people, products and markets, infrastructure and data, and governance.
Why education requires particular care
Education technology does more than process ordinary administrative information. It can hold student identities, grades, learning difficulties, disability information, attendance records, behaviour patterns and interactions between learners and teachers.
Increasingly, systems also create derived data: predictions, profiles, embeddings, risk scores, recommendations and behavioural analytics. This information may influence assessment, student support, course selection, admissions, progression and access to opportunities. Many of the people concerned are children or young adults.
Education platforms can also shape pedagogy. A system’s features determine what teachers can create, what learners can access, how collaboration takes place and which forms of assessment are easiest to implement.
When one provider controls identity, communication, cloud hosting, learning management, analytics and AI assistance, the issue is no longer simply which software the institution uses. The provider is beginning to shape the institution’s operating environment.
That is why digital sovereignty must be treated as an educational responsibility, not only as a technology policy.
Four shifts that can make sovereignty operational
Move from product purchasing to lifecycle procurement
Procurement is one of the places where sovereignty becomes real. Institutions frequently evaluate products according to price, functionality, usability and basic legal compliance. These factors remain important, but they are not enough.
Every significant EdTech procurement should address the complete lifecycle of the relationship, including the end of the contract.
An exit plan should be agreed before the product is adopted. It should explain how the institution can retrieve learning records, content, metadata, configurations and other necessary information in usable formats. It should also identify the cost, responsibilities and expected timetable for migration.
Providers should demonstrate interoperability rather than merely state that integrations are available. Institutions should be able to test APIs, data exports and compatibility with identity, learning-management and credential systems.
Contracts must also state clearly whether learner, teacher or institutional data can be used for model training, profiling, advertising or product development.
For high-impact AI applications — such as assessment, student guidance, tutoring or learning analytics—institutions need meaningful transparency, human oversight, incident-reporting processes and appropriate audit rights.
Shared templates and evaluation tools could make these requirements easier for institutions and smaller providers alike. Sovereignty should not become a compliance exercise that only the largest companies can afford.
Move from closed platforms to federated infrastructure
Europe does not need to replace several global monopolies with one large European education platform. A better objective is a federated education architecture: shared foundations and standards that allow different providers and public institutions to work together while preserving local choice.
Such an architecture could include:
- Trusted identity and consent services
- Portable learning records and credentials
- Open educational resources with consistent metadata
- Model-agnostic AI gateways
- Secure access to verified institutional knowledge
- Transparent monitoring and incident reporting
- Local, national, European and trusted commercial hosting options
This would allow institutions to combine services according to their needs and risk levels. A low-risk classroom utility may require different conditions from a national learner-identity service, assessment platform or AI tutoring system. The aim is not to eliminate providers. It is to create public-interest foundations on which many providers can compete. A useful principle is: public institutions should govern the rails, while vendors compete on the services.
Move from technology users to informed co-designers
Digital sovereignty cannot be achieved through infrastructure and contracts alone. Educators, institutional leaders, learners, procurement teams and technology staff need the capacity to understand and shape the systems they use. AI literacy should not be reduced to a short course on writing prompts. It should include the ability to ask:
- What data is being collected?
- What are the limits of this system?
- How can bias or inaccurate output be identified?
- When should human judgement override automation?
- How can educators request changes?
- Who is accountable when something goes wrong?
Educators should be included in needs analysis, product selection, testing and evaluation. They should not discover a new system only after a procurement decision has already been made.
Institutions also need specialist roles that connect technology, pedagogy, law, governance and data. Not every school or university can employ every type of expert, which makes shared competence centres and cross-institutional networks particularly important.
Move from pilot funding to sustainable services
Europe has many promising pilots. Far fewer become maintained, dependable services.
Digital infrastructure requires security updates, documentation, technical support, accessibility improvements, translation, testing, community management and long-term governance. These activities may attract less attention than launching a new application, but they determine whether an innovation can be trusted and used over time.
Open-source technology is not automatically sustainable. Without maintenance funding, responsible governance and institutional adoption, it can become another form of deferred technical debt. Digital infrastructure should therefore be treated as a continuing institutional responsibility—more like a library, network or building than a temporary innovation project.
Funding programmes should support operation, maintenance, skills and cooperation as well as new development.
Six questions every education organisation can ask
Before introducing a significant EdTech or AI service, institutions can begin with six questions:
- What learner, teacher and institutional data does the system collect, including profiles, analytics and other derived data?
- Can we export our information, content and configurations in usable and documented formats?
- Could we replace this component without replacing the entire digital environment?
- Can the system’s integrations, AI behaviour and high-impact decisions be independently examined?
- What educational, accessibility and inclusion benefits does the system provide—and how will we measure them?
- Who will maintain the service, pay for its continued operation and manage an eventual migration?
These questions will not solve every issue, but they can reveal where an institution has genuine choice and where dependency is beginning to accumulate.
An invitation to the EdTech community
The sovereignty discussion can create real value for EdTech providers, educators, institutions, researchers and public authorities. Members could work together on common evaluation criteria, procurement guidance, interoperability tests, reference architectures, professional development and cross-border pilots. EdTech providers can contribute open interfaces, portability tools, transparent data practices and technical components. Institutions can offer use cases and pilot environments. Educators can identify genuine needs and test whether proposed solutions improve teaching and learning. Researchers can provide evidence, evaluation methods and independent analysis. Public authorities can create coordinated procurement and sustainable funding conditions.
The questions for our community are therefore practical:
- What shared resource would help you most?
- Which dependency is currently limiting your work?
- What expertise, technology, evidence or pilot opportunity could you contribute?
Digital sovereignty will not be delivered by one regulation, one company or one institution. It is a collective capacity that must be built and maintained across the ecosystem. The objective is not to protect European education from competition. It is to protect its capacity to choose.
