The Microsoft outage showed who owns your uptime

Microsoft outage

The Microsoft outage that began around noon on the US East Coast on Thursday was still widespread on Friday morning, as Europe woke up. Email, Teams meetings and shared files were out for companies, banks, courts and governments in many countries. Microsoft says a cybersecurity breach did not cause it.

Table of contents
  1. More than ten hours and no sustained recovery
  2. Microsoft’s accounts of the cause do not match
  3. Where plan B was a phone call
  4. Credits exist, but customers have to claim them
  5. Europe’s sovereignty talk meets a real Microsoft outage
  6. Questions before the next procurement round

The cause is described in conflicting ways. The damage is documented: one vendor’s capacity problem stopped routine work in what Downdetector counts as 74 countries and territories, reached within about an hour.

More than ten hours and no sustained recovery

Microsoft first said the problem was in the Americas, then confirmed it was global. Downdetector logged close to 130,000 reports at the peak, a figure that understates the problem because most affected users never report. Services that depend on Exchange Online (Microsoft’s hosted email back end) were hit: Outlook, Teams, SharePoint, other Microsoft 365 apps and the Copilot assistant.

Microsoft’s own updates swung. Hours in, it reported progress from restarting infrastructure. Later it said its changes were “not taking as expected”, and that servers returned to service could become “unresponsive again once they receive traffic”.

Microsoft’s accounts of the cause do not match

Reuters quotes a service notice saying servers became unexpectedly unavailable and that the problem was not driven by a rise in traffic, and later cites “sustained traffic management imbalances”. AFP reports Microsoft saying part of its infrastructure was under elevated load across several regions. Microsoft’s incident record, MO1550821, lists Exchange Online, SharePoint Online and Defender XDR among the affected services and dates the start to about 1:25 pm Eastern, although the first, smaller failures may have come earlier in the day. Microsoft says it paused deployments, freed memory, rerouted traffic and rebooted servers.

Read also: Starshield: Ukraine’s drone defence runs on a network Europe does not own

Australia’s home affairs and cybersecurity minister, Tony Burke, told a press conference that Microsoft believed it was not a cyberattack, describing an update that “has gone awry”. Microsoft gave no date for full restoration.

Where plan B was a phone call

Some US office workers shrugged and used Slack, the phone or a walk down the corridor. Others had no such option. City governments across the US told residents to call, because email failed. In Australia, courts and schools were among those hit. Longreach School of the Air in Queensland suspended live online lessons and digital communications “until further notice”.

New Zealand’s foreign ministry said its ability to receive email was affected, and that included its embassies and consulates. In Hong Kong, bankers at Morgan Stanley and Goldman Sachs said their work was disrupted. IndiGo and Akasa Air in India and CemAir in South Africa reported disruption to email support, and none reported flight disruption.

Credits exist, but customers have to claim them

Business customers have a contractual route to money back. For Exchange Online, Microsoft’s terms set three tiers: a 25% credit if monthly uptime falls below 99.9%, 50% below 99% and 100% below 95%. The percentages apply to the fees of the affected services, not the whole bill, and uptime is measured in user-minutes, so a global outage does not automatically put every organisation in the same tier.

Posts claiming that seven hours of downtime means an automatic 50% credit oversimplify. In a 30-day month of 43,200 minutes, about 43 minutes of downtime for every user breaches 99.9%, and about seven hours 12 minutes breaches 99%. If only some users were hit, or the service worked intermittently, the arithmetic changes.

Nothing is paid out unprompted. Customers who bought directly from Microsoft open a support request in the Microsoft 365 admin centre, ask for an SLA credit review and quote the incident ID: MO1550821 for the wider Microsoft 365 incident, EX1550543 for Exchange Online. Customers who buy through a reseller or Cloud Solution Provider must ask that provider to file the claim. For services other than Azure, Microsoft’s terms set the deadline at one month after the end of the billing month in which the incident happened. Personal Outlook.com and Microsoft 365 Personal and Family accounts are generally not covered.

A credit is a discount on the subscription. It does not cover the hours that courts, schools and ministries lost.

Read also: The Digital Omnibus Follows Big Tech’s Wish List Closely

Europe’s sovereignty talk meets a real Microsoft outage

The outage lands in a year when European governments say they want less dependence on US technology. French Prime Minister Sébastien Lecornu said earlier this year that officials would drop Zoom and other US-owned video tools for French-owned software, partly to cut “dependencies on non-European actors”. Neil Shearing, group chief economist at Capital Economics, has described Europe’s internet as essentially a US construct with no European rival.

Technology journalist Jacob Ward called the outage “a wake up call about just how vulnerable we can be”. Downdetector’s map showed users hit across parts of Europe, and the outage was still widespread when the continent woke on Friday.

This is not the first time. When Azure failed on 29 October 2025, the Dutch rail operator NS lost its travel planner and ticket machines. Governments have started to move: the European Commission has launched a €180 million tender for digitally sovereign clouds, the German state of Schleswig-Holstein is shifting its administration away from Microsoft products, and Denmark’s digital ministry, Copenhagen and Aarhus have announced similar steps.

Cloud providers will say outages are rare and that most workers coped. Many did. Those who could not were courts, schools, ministries and city halls, whose work cannot wait for an inbox to return.

Questions before the next procurement round

Any administration that runs email, calendars and meetings on one vendor shares this exposure, whatever the contract says about sovereignty. How long can the office work without the vendor, and which fallback channel is written down and tested? Who pays when an update goes wrong?

An outage of this length is the cheapest version of this test. Nobody has to wait for a longer one to write the fallback plan.

Part of our guide to Europe’s digital dependence: Europe Runs on Computers It Does Not Own

BASIS OF ANALYSIS

Basis of analysis. The analysis rests on ten sources, only two of them not media reports (a case catalogue prepared for public bodies and Microsoft’s own training material on SLAs), and Microsoft’s statements are reproduced through media. The cause of the outage and the full extent of the impact in Europe could not be confirmed.

Independent journalism

ExtraFokus is ad-free, and that costs money

Be one of the first 100 members of ExtraFokus.eu: €8/month.

No comments

Your email address will not be published. Required fields are marked *

Newsletter

Subscribe to our newsletter to get our latest articles right away!

Support ExtraFokus

Be one of the first 100 members of ExtraFokus.eu. No ads, no media group, just readers like you.

Newsletter

Subscribe to our newsletter to get our latest articles right away!

Support ExtraFokus

Be one of the first 100 members of ExtraFokus.eu. No ads, no media group, just readers like you.

Become a Citizen Journalist

Seventeen free lessons in finding stories, checking facts, writing and treating people fairly.