European .IA initiative: sovereign governance, not public policy yet

Article created 8 October 2026 · Publication analysed: 6 October 2026 · Source: European Commission

The European Commission has registered a citizens’ initiative for the “Creation of a European Digital Space (.IA) for Sovereign AI Governance”. It concerns high-trust digital spaces, testing environments, and the .IA domain name. It is a political signal, not an applicable rule or proof of sovereignty.

1. What was registered

The Commission declared the initiative formally admissible. Its organisers can open a 12-month collection period. A Commission response is only required after at least one million valid statements of support across at least seven Member States. Registration therefore does not approve the proposal, create a European domain, or commit the Commission to negotiate with ISO or ICANN.

2. A domain is not a trust boundary

An internet suffix does not control hosting, data location, administrator access, or the software supply chain. For an organisation in Belgium or France, a high-trust space must translate into evidence: an inventory of data and models, processing countries, permissions, exportable logs, audit, reversibility, and contractual responsibilities.

3. Identity, compliance, and infrastructure remain distinct

eIDAS can govern certain identity evidence; the AI Act structures obligations by use case; and the GDPR addresses personal data among other matters. They do not replace architecture and operations. In an agent connected to Odoo or an ERP, strong identity must not automatically grant authority to approve a payment or change master data.

4. The control to exercise now

The initiative is worth monitoring, but teams should not wait for its outcome to control flows. Each use case needs to document data, model, runtime, tools, identities, country, logs, support, and exit.

Read the official announcement