Microsoft 365 fully restored its services early on Saturday, November 14, after a global outage that began on Thursday the 12th and left companies, hospitals and governments in 74 countries without Outlook email, Teams or Copilot, according to Xataka. The outage began at 5:17 p.m. Spain time (1:17 p.m. in Chile) and the company announced full recovery shortly after 7:09 a.m. Spain time on Saturday (3:09 a.m. in Chile): about 38 hours in total.
Microsoft attributed the failure to a change in its network infrastructure, ruled out a cyberattack and promised a full report once its review is complete. It is one of the longest cloud outages in recent years. This is what is known about the Microsoft 365 outage and how it compares with the CrowdStrike outage in 2024 and the Amazon Web Services outage in 2025.
View the data as a table
| Approximate duration | 38 hours |
| Countries with Downdetector reports | 74 |
| Main services affected | 3 |
Timeline of the Microsoft 365 outage
The problems began in Exchange Online, the service behind Outlook email, and during Thursday afternoon and night spread to Teams, with its video calls, and to the Copilot artificial intelligence assistant. According to FayerWayer, reports reached 74 countries and territories in about an hour. On Friday morning there were improvements in several regions, but access problems and email delays continued.
- Thu 12, 1:17 p.m.Exchange Online access failuresFirst email access problems; the failures start in the United States.
- Thu 12, afternoon and nightSpreads to Teams and CopilotReports in 74 countries and territories.
- Fri 13, morningPartial improvementsMicrosoft reports progress; Latin American banks report problems in their apps.
- Fri 13, 1:40 p.m.Widespread but incomplete recoverySome customers and regions are still affected.
- Sat 14, 3:09 a.m.Full recoveryMicrosoft 365 Status announces it on X.
Source: Xataka, with Spain times converted to Chile time (four hours earlier). The duration is an approximate calculation between the first report and the full recovery announcement; some customers got their service back earlier.
What caused the Microsoft 365 outage
According to Microsoft's messages reported by Xataka, the problem began during an earlier update to its network capacity and surfaced when a subsequent change was made. The company began rolling back a configuration change and restored service gradually, validating each stage, which according to Microsoft itself lengthened the recovery. It ruled out a cyberattack or a security problem.
Microsoft has not yet published its final technical report, so the exact cause may be clarified in the coming days. It has also not given a figure for affected users.
Who the Microsoft outage affected
Among those affected cited by Xataka are the shipping company Maersk, the regional government of Andalusia, British National Health Service hospitals (Guy's and St Thomas'), Los Angeles authorities and Stanford University; according to Reuters, some companies went back to using fax. FayerWayer also mentions rail systems, airports and courts.
In Latin America, several banks and mobile banking apps recorded failures on Downdetector on Friday the 13th, which FayerWayer attributed to a "direct side effect" of the outage; the institutions asked customers to use alternative channels or ATMs. The outlet did not identify the banks or the countries.
| Sector | Affected organizations cited |
|---|---|
| Logistics | Maersk |
| Health | Guy's and St Thomas' hospitals (NHS, United Kingdom) |
| Government | Regional government of Andalusia; Los Angeles authorities |
| Education | Stanford University |
| Banking | Banks and mobile banking apps in Latin America (unnamed) |
Source: Xataka; FayerWayer; FayerWayer.
The Microsoft outage compared with other major cloud outages
The November 2026 outage is longer than the other major recent disruptions. The Amazon Web Services outage in its Northern Virginia region, in October 2025, lasted about 14 and a half hours, from 11:48 p.m. on October 19 to 2:20 p.m. on the 20th (Pacific time), due to a fault in DynamoDB's automated DNS management, according to the AWS report. In January 2026, another Microsoft 365 outage lasted almost 10 hours, according to The Register.
| Date | Company | What failed | Scope |
|---|---|---|---|
| Jul 19, 2024 | CrowdStrike / Microsoft | Faulty update of the Falcon antivirus on Windows | 8.5 million Windows devices, according to Microsoft |
| Oct 19-20, 2025 | Amazon Web Services | DynamoDB DNS in the us-east-1 region | About 14.5 hours, according to AWS |
| Jan 2026 | Microsoft 365 | Exchange Online and other services | Nearly 10 hours, according to The Register |
| Nov 12-14, 2026 | Microsoft 365 | Network infrastructure change | About 38 hours, 74 countries, according to Xataka |
The outages show how much the economy depends on a few providers: according to the European Commission, cited by Xataka, Amazon (AWS), Microsoft (Azure) and Google Cloud account for more than 70% of the European cloud market. In Chile, the cloud is physically arriving in the country with new data centers, such as the Amazon Web Services region, and the Latin American Artificial Intelligence Index measures how much computing capacity each country has.
Frequently asked questions
Is Outlook working again?
Yes. Microsoft announced the full recovery of Microsoft 365 early on Saturday, November 14 (shortly after 3:09 a.m. Chile time). If you still have problems, check the Microsoft 365 status page.
Why did Microsoft 365 go down?
Microsoft attributes it to a change in its network infrastructure following a capacity update. It ruled out a cyberattack and will publish a full report.
How long did the outage last?
About 38 hours between the first report, on Thursday the 12th at 1:17 p.m. Chile time, and full recovery on Saturday the 14th.
Were emails lost?
Microsoft has not reported any data loss; reports mention delays in email delivery during the outage.
