SovereignEU Plan a deployment
Menu

Public-private collaboration across Europe

European Alliance for Open-Source Sovereign AI.

Research. Open technology. Professional delivery. Customer control.

SovereignEU brings together public and private organisations across Europe and collaborates with public universities, research centres, local infrastructure providers and European companies advancing open-source software and AI.

About the alliance Explore SovereignOS

The alliance offer

Research, open technology and accountable delivery.

SovereignEU works across public organisations, research and European industry to turn shared research and development into deployable organisational capabilities.

Discover how the alliance works

Research and development

Collaborative research, evaluation and engineering with public universities, research centres and European technology teams.

SovereignOS and applications

A flagship operating-system Roadmap integrating organisational applications, governed intelligence, core services and trust controls.

Professional services

Assessment, integration, migration, deployment, hardening, governance, training and support with explicit responsibilities and access boundaries.

Flagship operating system · Roadmap

One AI-native operating system for sovereign work.

SovereignOS is the planned workstation environment for SovereignEU applications, organisation-controlled AI and customer-defined operations. It is designed as an alternative to proprietary desktop operating systems for validated hardware and workflows.

Explore SovereignOS
SovereignOS Regional administration · Customer boundary Local control
AI intentPrepare the mobility committee brief from approved project records Super K
Office ApplicationsDraft · source-linked

REGIONAL TRANSPORT · 23 AUGUST 2026

Decision brief: shared mobility procurement

Three open decisions have been assembled from the approved project register, procurement policy and committee record.

Recommended next stepConfirm accessibility criteria before tender release.

6 approved sources · 2 passages need review

Organisation identityLocal model routeSigned update policyReviewable activity record
Applications and governed AI in one operating environment.Roadmap interface concept. It illustrates the intended relationship among applications, context, models, permissions and approval; it does not represent a currently available release.

Inside SovereignOS

One connected portfolio. Four clear responsibilities.

The alliance’s seventeen Roadmap products sit inside one operating-system direction—from daily organisational work to shared intelligence, core services and trust. Applications are integrated by default while documented interfaces and export paths preserve portability.

Applications

Organisational applications

Everyday tools for content, knowledge, coordination and communication.

Intelligence

Intelligence and automation

Local document understanding, assistants and model adaptation.

Foundation

Data and runtime foundation

Shared data, model-serving and identity foundations.

Trust

AI, cryptography and operational assurance

AI controls, defense and cryptographic transition.

Product descriptions and planned replacement scope are available in the complete catalogue.

Explore the complete product family

Sovereign Intelligence · Roadmap

Context that outlives the prompt.

Sovereign Intelligence is the planned intelligence layer across SovereignEU applications. It combines permission-aware temporal retrieval, customer-selected models and policy-controlled tools so assistants can answer and act from relevant, source-linked evidence.

Persistent organisational context. Bounded, auditable prompts. A model’s input remains finite; the graph preserves governed memory and assembles only the relevant evidence for each task.

Explore Sovereign Intelligence
  1. 01
    ConnectApproved records, work, policies and history
  2. 02
    RetrieveA bounded, permission-checked subgraph
  3. 03
    ReasonSpecialist agents investigate source-linked evidence
  4. 04
    ActAuthorised tools operate inside policy boundaries
  5. 05
    EvidencePermitted sources, actions, approvals and outcomes

Solutions by organisational function · Roadmap

One context layer. Different responsibilities.

Each function keeps its own permissions, records and workflows while sharing only approved knowledge and evidence through SovereignOS.

Explore all solutions

Deliberate scope: this roadmap focuses on institutional operations and public-service delivery—not marketing or sales.

Defining Roadmap capability

Shared context can support governed improvement.

Customer-authorised events and outcomes can be related to workflows, versions and releases in the temporal graph. Local agents may diagnose friction and prepare candidate changes under customer-defined evaluation, approval, versioning and rollback controls.

Explore the Improvement Loop

Governed automation

Automatic analysis. Controlled release.

Customer policy defines authorised signals and determines which predefined low-risk changes may follow a pre-authorised path and which always require accountable human approval.

Review the trust model

European regulation by design · Roadmap

AI Act and GDPR controls across the complete system.

European requirements shape the Roadmap across applications, models, data, devices and operations. SovereignOS is being designed to give accountable organisations the controls and technical evidence needed to govern AI inside their own boundary.

Entire system European governance controls Applications · models · data · operations
01

Data protection

Purpose, permissions, minimisation, retention, location and export controls.

02

AI lifecycle governance

System inventories, data and model provenance, risk controls and evaluations.

03

Human oversight

Accountable approvals, intervention, pause, rollback and escalation paths.

04

Technical evidence

Logs, documentation, decisions, incidents and reviewable audit exports.

These planned controls can support organisational work under the GDPR, EU AI Act and applicable European security and accessibility obligations. Legal compliance remains specific to the released version, deployment, use case and organisation.
Explore governance and assurance

A verifiable definition

Sovereignty is more than a hosting location.

A credible deployment defines control across data, operation, technology, jurisdiction, organisational knowledge and exit.

Data

Where can data, prompts, outputs, logs and backups go?

Operational

Can your organisation keep operating without a mandatory vendor cloud?

Technology

Can components be inspected, integrated, adapted and replaced?

Jurisdiction

Which entities, staff and subprocessors can access the service?

Knowledge

Can your teams understand, operate and improve what was delivered?

Exit

Can data, policies, models and workflows move in usable formats?

Understand the full workflow

A measured path to operation

Begin with one bounded use case.

Define success before implementation, prove the operational boundary and transfer the knowledge needed to run it.

  1. 01
    Assess

    Map the workflow, data, dependencies and required control boundary.

  2. 02
    Define

    Agree quality, security, usability, operational and exit criteria.

  3. 03
    Deploy

    Introduce a bounded pilot alongside existing systems where needed.

  4. 04
    Verify and transfer

    Test the result, document operation and equip the customer team.

A practical first step

Bring your sovereign AI challenge to the alliance.

Start with a workflow, research question or delivery requirement and define the applications, data, models, controls and operating boundary involved.

Start a conversation