Blue technical illustration of Canadian and European energy, data and communications infrastructure connected across the Atlantic.

Insights

Digital Sovereignty: What Canada and Europe Can Build Together

Manuel Bohé
Manuel BohéNov 17, 2026 · CEO

Digital sovereignty starts with operational choices. How Canadian and European infrastructure operators can modernise essential systems, retain control and bring useful AI into service.

Consider a routine technology renewal at a Canadian infrastructure operator. A familiar supplier offers a more integrated platform, new AI capabilities and a simpler support model. The proposal looks attractive. It also raises questions: which operational data would leave the organisation, which functions would depend on an external connection, and how difficult would another change of direction become?

The decision has to reconcile several legitimate priorities. Operations needs continuity. Security needs control. Employees expect useful tools, and management expects investment to improve performance. Digital sovereignty becomes meaningful where those expectations meet.

Geopolitics adds urgency. Shared Services Canada explicitly identifies foreign supplier dependence, vendor lock-in and geopolitical tensions as risks to control over government systems and sensitive data (Shared Services Canada, 2026b). For a critical infrastructure operator, the practical question is how much freedom remains to operate, adapt and recover essential systems when circumstances change.

The political ambition now reflects that shared challenge. European Commission President Ursula von der Leyen has called for a closer partnership with Canada, encompassing areas such as AI, energy and security (European Commission, 2026). Prime Minister Mark Carney has advocated the “deepest possible partnership” between Canada and the EU, grounded in complementary strengths and greater sovereignty for both (Prime Minister of Canada, 2026). At the Montréal summit on 29–30 October 2026, the two sides launched the Alliance for the Future to advance that cooperation (European Council, 2026).

For infrastructure operators, the opportunity is to turn that ambition into practical work: maintaining essential services, making technology easier to adapt and bringing useful AI into environments where control matters. Shared projects could begin with a specific operational requirement and bring together operators, engineers and technology partners around a solution that each can support. Existing deployments on both sides of the Atlantic provide starting points.

Start with the function that must remain available

For a team responsible for essential services, a technology transition has to earn its place. Existing applications, working practices and supplier relationships all have operational value. A credible alternative needs a clear purpose and a manageable path into service.

Canada’s approach to secure government communications illustrates how a defined requirement can shape the operating model: Shared Services Canada describes encrypted calls and messaging using SecuSUITE, with communications and data hosted in Government of Canada data centres (Shared Services Canada, 2026a).

In an interview with Saunders (2026), Austria’s armed forces describe their move to LibreOffice as an investment in functioning without public infrastructure or external data processing, backed by substantial preparation, training, testing, support and funded development to meet their interoperability needs. The operational requirement gives the transition its purpose; the implementation effort makes it sustainable.

The two services address different needs, but they open a useful conversation about where control belongs. For an infrastructure operator, a maintenance team may need technical documentation during a connectivity failure; an incident team may need access to procedures while parts of the corporate environment are isolated. Starting with the required function makes it possible to define what independence should achieve and where an existing dependency remains acceptable.

It also changes the investment discussion. Support, integration and staff capability belong in the business case from the beginning. The alternative has to work on an ordinary Tuesday as well as during the disruption that justified it.

The same question reaches into the substation

In operational technology, the consequences of a platform decision extend to equipment lifecycles, engineering tools and future supplier choice. An operator considering modernisation needs to know how much can change without reopening the entire technical design.

RTE’s work with SEAPATH offers a concrete example: the French transmission system operator was running the open-source virtualisation platform in production across six substations, enabling applications to run on standard server hardware, while extension to critical protection workloads remained a separate research and development effort (LF Energy, n.d.-b). For an operator planning future investment, separating software functions from dedicated equipment creates room to reconsider how the system is built and maintained.

This creates an opportunity for joint engineering work around a shared platform. A Canadian utility and its partners could assess which functions fit their own requirements, contribute improvements and compare validation results with other operators. The engineering questions remain demanding: real-time performance, compatibility, secure maintenance and long-term support must all be demonstrated for the intended application.

TenneT Netherlands provides another operational example: its ReFlow platform combines the open-source PowSyBl engine with an architecture designed to accommodate other calculation engines, an existing on-premises platform and engineers embedded in the operational business, with TenneT reporting at least tenfold performance improvements over its previous solution and a first production release after five months (LF Energy, 2026). The combination matters: the result brings together open software, in-house infrastructure and operational expertise.

AI remains part of the modernisation path

Any alternative technology strategy now has to answer a reasonable question: will it leave employees and engineers with access to useful AI? The answer depends on decisions about the application, the model and the infrastructure on which it runs.

Baden-Württemberg operates its F13 AI assistant in the data centre of its state-owned IT provider, BITBW, and has made a model-agnostic version available for reuse as F13-OS (Government of Baden-Württemberg, 2025). This opens up choices over both the models an organisation uses and the environment in which it operates them.

The opportunity extends beyond document assistance. OpenSTEF provides open-source machine-learning pipelines for short-term energy forecasting, with operational use at Dutch network operator Alliander documented by LF Energy (LF Energy, n.d.-a).

For an operator considering AI, the next step is to evaluate a defined task under the conditions in which it would actually be used. The assessment should establish whether the application produces useful results, which data it needs, where that data is processed and who maintains the service. It should also test how the task would be carried out if the AI were unavailable.

Own infrastructure and managed regional services can both be considered. The appropriate choice needs to account for application quality, integration, support, cost and the dependencies that remain.

Preserve control as intelligence moves closer to operations

The closer an AI application comes to a physical process, the more important its authority becomes. A maintenance recommendation and an automatic change to a control setting require different levels of assurance.

There is already a common foundation for this work: international guidance co-authored by Canada’s Cyber Centre and European partner agencies recommends controlled data flows out of OT and conventional or manual fallback, illustrates read-only predictive maintenance, and highlights risks from centralising sensitive OT data and relying on LLMs for safety decisions (CISA et al., 2025). Canada’s guidance on edge AI recommends prioritising local processing of sensitive data where feasible, alongside protections against tampering, trusted software and models, and independent safety mechanisms (Canadian Centre for Cyber Security, 2026).

A sensible starting point is an application that analyses approved data and returns advice, with operational changes subject to an authorised process. Any proposed control function needs a separate engineering and functional-safety assessment, with defined operating limits and tested fallback.

Build cooperation around an operational result

The strongest starting point for cooperation is an operational requirement that partners can work on together. Select an important function, identify the dependency that constrains it and evaluate an alternative under realistic conditions. A useful pilot should produce evidence of performance, a tested fallback and a clear account of the effort required to sustain the service.

For a Canadian infrastructure operator, a useful outcome is a system that performs its task well, can be supported by people with clear responsibilities and leaves credible options for its future. An assessment of sovereignty should therefore examine continuity alongside access rights, applicable legal obligations, control over updates and data, and the practical ability to change suppliers. Local hosting and open-source components are choices to evaluate within that assessment; neither alone establishes control over the whole service.

Concepture is an independent, vendor-neutral security consultancy. We work with infrastructure operators to design security architectures around their operational needs and the threats they face.

In Canada, we seek to exchange experience and develop approaches with local operators and technology partners, drawing on our work in Europe.

We also help Canadian technology companies identify relevant partners and opportunities in Germany and the wider European Union.

References

Canadian Centre for Cyber Security (2026) ‘Securely deploying AI at the network edge – ITSP.80.101’, 14 July. Available at: View publication (Accessed: 17 November 2026).

CISA (Cybersecurity and Infrastructure Security Agency), Australian Signals Directorate’s Australian Cyber Security Centre, National Security Agency’s Artificial Intelligence Security Center, Federal Bureau of Investigation, Canadian Centre for Cyber Security, German Federal Office for Information Security, Netherlands National Cyber Security Centre, New Zealand National Cyber Security Centre and United Kingdom National Cyber Security Centre (2025) ‘Principles for the secure integration of Artificial Intelligence in Operational Technology’, 4 December. Available at: View publication (Accessed: 17 November 2026).

European Commission (2026) ‘Gipfeltreffen EU-Kanada 2026: Politische Führungsspitzen markieren historischen Moment in den bilateralen Beziehungen’, 31 October. Available at: View publication (Accessed: 17 November 2026).

European Council (2026) ‘EU-Canada summit, 29–30 October 2026’. Available at: View publication (Accessed: 17 November 2026).

Government of Baden-Württemberg (2025) ‘KI-Assistenz F13 wird zur Open-Source-Software’, 23 July. Available at: View publication (Accessed: 17 November 2026).

LF Energy (2026) ‘Grid Security at Scale: How TenneT Built a 10× Faster Analysis Platform on PowSyBl’. Updated: 2 April. Available at: View publication (Accessed: 17 November 2026).

LF Energy (n.d.-a) ‘OpenSTEF’. Available at: View publication (Accessed: 17 November 2026).

LF Energy (n.d.-b) ‘The SEAPATH Roadmap Workshop 2026 – A Growing Ecosystem Shapes the Future of Virtual Substations’. Available at: View publication (Accessed: 17 November 2026).

Prime Minister of Canada (2026) ‘Prime Minister Carney deepens ties with trusted allies in visit to France and the United Kingdom’, 17 September. Available at: View publication (Accessed: 17 November 2026).

Saunders, M. (2026) ‘Why and how the Austrian Military moved to LibreOffice (Part 2)’, TDF Community Blog, 14 August. Available at: View publication (Accessed: 17 November 2026).

Shared Services Canada (2026a) ‘Digital sovereignty and secure communications’. Available at: View publication (Accessed: 17 November 2026).

Shared Services Canada (2026b) ‘Shared Services Canada 2026–27 Departmental Plan’. Available at: View publication (Accessed: 17 November 2026).