Opinion: Sovereignty Did Not Answer the Phone. Plans A, B and C Did.
Opinion. After the Microsoft outage, Europe called again for digital sovereignty. The organisations that kept working had something simpler: three tested fallback plans.
Opinion. This piece gives the view of the InekIA editorial team. The facts it relies on are set out, with their sources, in our factual article: Microsoft Outage: A 24-Hour Failure and European Continuity.
For almost 24 hours from 12 November 2026, Outlook, Teams and the rest of Microsoft 365 were unavailable in at least 74 countries and territories. Within hours, European lawmakers renewed their call for local cloud services. That call has been made for years. Meanwhile Euronext, the UK government, Scotland's prosecution service and the NHS were all disrupted by the failure of the same provider.
Digital sovereignty is a wish, not a plan. European providers are reported to hold about 15% of the European cloud market, a figure from a secondary source that no official statistic confirms. No risk committee can wait for that share to change.
The useful question is narrower: what do you do in the first hour, on the first day and in the first week?
Plan A: reduce dependence where it is real
Changing suppliers does not create independence if the new one shares identity, infrastructure or security with the old one, as Control Risks notes. So start with what has to survive a provider failure: identity, an emergency communication channel, and the continuity plan itself. If the plan lives in SharePoint, there is no plan.
Plan B: a degraded mode that has been rehearsed
The organisations that coped had one. Euronext moved trade communications to the telephone. East Lothian Council activated its continuity plan. In one Gulf company, SAP kept running and business still stopped, because approvals needed email. The weak point was the process, not the cloud. Write a backup approval rule and test it more than once a year.
Plan C: the outage that lasts days
Telephone and fax do not cover a week. Improvised channels such as Gmail or WhatsApp raise record-keeping problems. A pre-agreed alternative, with its compliance questions settled in advance, answers the question organisations under DORA and NIS2 now face: did continuity plans and third-party arrangements hold?
Who pays for what
Risk owners need to fund all three plans. Procurement needs to know which dependencies are shared before signing a second contract. Compliance needs to check that the fallback channel keeps records.
What we do not know
Microsoft has not published its root-cause report, and there is no official figure for European losses. If the report shows an architectural weakness, Plan A becomes more urgent. Either way, a company with three tested plans is better protected than one waiting for a sovereign cloud.
— InekIA editorial team