Quick Answer: The Microsoft 365 outage was fully resolved on Saturday, November 14, 2026, after a multi-hour disruption that began midday Thursday. Microsoft said an issue during an earlier network capacity upgrade caused it, not a cyberattack or AI agents. Its reach was wide because many governments, courts, banks, and clinics rely on the same platform.
If your inbox went quiet on Thursday, you weren't alone. The Microsoft 365 outage started around midday in the U.S., rolled into Friday, and wasn't fully cleared until Saturday. Downdetector logged reports from users in 74 countries and territories within the first hour. The story isn't just that one company had a bad day. It's what happens when one provider sits under your email, meetings, and sign-ins at the same time.
What Actually Broke, and What Microsoft Said Caused It
Users reported that Outlook, Teams, and the Copilot AI assistant were all interrupted. For many people, the visible symptom was simple: they couldn't send or receive email, or their inbox wouldn't load at all. Microsoft's engineers traced part of the problem to Exchange Online mailboxes. The company said a failure left a number of its servers "unexpectedly unavailable" during automated checks.
The root explanation Microsoft gave was an issue during a network capacity upgrade that followed a recent network change. Microsoft told the Associated Press that the incident was "not related to cyber" and that there was "no indication" of malicious activity.
Scale explains why a single fault travels so far. Microsoft says more than 430 million people use Microsoft 365 apps, and it reported more than 450 million paid commercial subscriptions in January. When a platform that size fails in one layer, the blast radius is measured in governments and hospitals, not just companies.
Here's the part most coverage skips. A change that's entirely benign, with no attacker involved, can still take down services for hours. Security teams are trained to ask "was this an intrusion?" first. When the answer is no, many organizations relax, but their exposure hasn't changed at all.
Why the Dependency Chain Reached Courts and Pharmacies
The disruption didn't stay inside tech departments. In Manchester, Judge Maurice Greene sent jurors home because delays illustrated the risk of relying on technology in court proceedings. British police said they couldn't receive crime reports online and asked people to phone urgent requests in. In Colorado, the Arapahoe County sheriff's office said online bond payments were affected.
Healthcare felt it too. The National Health Service was affected in ways that threatened to stop doctors from emailing prescriptions to pharmacies. The National Pharmacy Association warned customers that waits could be longer and some prescriptions delayed.
Governments in Singapore, Japan, Greece, Norway, and Australia reported that email for certain ministries or agencies was disrupted. Private businesses felt it just as sharply. A Hong Kong aluminum company couldn't receive updates from its U.S. customers through Outlook, and some banks in Hong Kong reported disruptions.
Even AI platforms were downstream. OpenAI and Anthropic both reported problems connecting with Microsoft 365 tools. As more products call other companies' software through APIs, one provider's outage can break a chain you didn't know existed.
Academics are already raising the policy question. Jeremy Singer, a professor at the University of Glasgow's School of Computing Science, said the outage "serves to emphasize the dependence we have built on one company's digital infrastructure." That's the right framing for your planning, even if you never think about geopolitics.
The Counter-Intuitive Problem: Recovery Is Slow and Retries Hurt
Most advice after an outage is to "try again once it's back." That's incomplete. Microsoft said it restored services gradually, and that the gradual approach took longer. It also said it was addressing "remaining impacted customers and regions" even as others came back online.
Partial recovery creates its own failure mode. When a service returns for some tenants or regions and not others, users who got errors hit send, resubmit, or reconnect in bulk. Queued messages can arrive late, duplicated, or out of order. Calendar invites and shared documents can show stale states for hours after the banner says "resolved."
This is an inference from how partial restorations usually behave, not something Microsoft confirmed about its own queues. Still, it's a reasonable assumption for planning. Treat "restored" in the status page as the start of your verification, not the end. Check that a test message arrives, that your calendar syncs, and that shared files open before you tell staff to resume normal work.
Here's where most guides go wrong: they assume the outage has a clean start and finish. For a multi-region incident, your recovery window may be a day longer than the headline suggests.
Continuity Options Compared
No single fallback covers everything. The right choice depends on what your team actually needs during downtime: reading mail, sending messages, joining meetings, or completing regulated tasks.
| Option | What it keeps working | Main limitation | Best for |
|---|---|---|---|
| Outlook cached mode | Reading and composing recent mail offline | Doesn't send until the connection returns; no new inbound mail | Individual staff who need to draft and read during outages |
| Phone and SMS escalation tree | Urgent contact with staff, clients, and vendors | Manual, slow for bulk comms, needs current contact data | Incident coordination and urgent notices |
| Secondary email relay or alternate domain | Outbound notices and some inbound mail | Requires setup, DNS changes, and regular testing | Customer-facing notices and regulated communications |
| Microsoft 365 Service Health alerts plus Downdetector | Early warning and public signal of an outage | Doesn't restore any service; it only speeds your response | Detection and communications timing |
Note that Microsoft's published service level agreement for most Microsoft 365 services commits to 99.9% uptime. That allows about 8.8 hours of downtime in a year. Service credits are generally the remedy, and they won't cover your lost billable hours or a delayed court filing. Read the SLA before you assume the vendor carries your downtime costs.
A Practical Continuity Plan You Can Finish This Month
Work through these steps in order. Each one is designed to reduce a specific failure, not to give you a generic checklist.
- Map the processes that can't pause. List the three to five workflows where delay causes real harm, such as prescriptions, payments, court deadlines, or client intake. For each, note whether it depends on Exchange Online, Teams, or sign-in.
- Create an out-of-band contact path. Store phone numbers for key staff outside Microsoft 365. The UK police response shows why: when reporting systems went down, people were told to phone instead.
- Turn on alerts before you need them. Subscribe to Microsoft 365 Service Health notifications for your tenant and watch Downdetector for early signals. Early detection gives you more time to act.
- Enable Outlook cached mode on all managed devices. It won't fix a server outage, but it keeps people productive while sending and syncing wait.
- Document manual fallbacks for regulated tasks. Write one-page instructions for the workflow, including who can authorize exceptions and how records get backfilled later.
- Test the plan. Run a 30-minute tabletop exercise with one scenario: "email and sign-in are down for six hours." Time how long it takes to notify staff and resume one critical task.
Most teams find step six reveals gaps they'd never have spotted from a document alone.
Frequently Asked Questions
What is the Microsoft 365 outage?
The Microsoft 365 outage was a multi-hour disruption that began around midday Thursday, November 12, 2026, and was declared fully restored on Saturday. It affected Outlook, Teams, and Copilot, with Exchange Online mailboxes a key part of the failure. Microsoft attributed it to an issue during a network capacity upgrade.
Was the Microsoft 365 outage caused by a cyberattack?
No, according to Microsoft. The company said the incident was "not related to cyber" or AI agents and that there was "no indication" of malicious activity. A non-malicious cause still exposes organizations to downtime, so the continuity steps above still apply.
How do I keep working during a Microsoft 365 outage?
Enable Outlook cached mode so you can read and draft mail offline, and set up a phone or SMS contact list for urgent communication. For critical workflows, write down manual fallbacks and who can approve them. Then check that messages and calendars have synced before you resume normal work.
Why did one Microsoft 365 outage affect so many countries?
Microsoft 365 is used by hundreds of millions of people, and many organizations rely on it for email, meetings, and identity. Downdetector received reports from 74 countries and territories within an hour. When one provider underpins that many functions, a single fault spreads across borders quickly.
When should I consider a backup email system?
Consider one when a few hours of email loss would stop a revenue process, a legal deadline, or patient care. For a small team with low email dependence, cached mode and a phone tree may be enough. Weigh the setup cost against the realistic cost of downtime, not the worst case.
Plan for the Next Outage Before It Happens
The most useful takeaway is simple: a Microsoft 365 outage is not a question of if, but of which of your processes can keep running when it hits. Pick one workflow that can't pause, write a manual fallback for it, and run a six-hour drill this week. Then send this article to whoever on your team owns email, sign-in, or client communication, and ask them to name the one process they'd fix first.