Microsoft Outlook and Teams Down: Outage Passes 15 Hours

Microsoft Outlook, Teams, Copilot and other Microsoft 365 services have been failing for users around the world since Thursday, Nov. 12, and by Friday, Nov. 13, the outage had run past 15 hours, according to CNA. Microsoft said in status updates that it sees early signs of stabilisation, but service capacity has not returned to normal and recovery remains uneven.

Microsoft said it is not aware of any malicious activity. It attributes the problem to a large number of its servers becoming unexpectedly unavailable.

What is not working and who is affected

Many users could not open their inboxes or send and receive email. In Teams, people had trouble sending messages and opening chat windows. Microsoft first said it was investigating reports in the Americas, then confirmed that Microsoft 365 connectivity could be affected for users in any region.

US users reported trouble by about 1 p.m. Eastern on Nov. 12, according to the AP. Government bodies in Australia, Japan and the United States were affected, and ABC reported that Australia’s spy agency and federal police were among the organizations caught up. The US House of Representatives said the outage hit its users and that its engineers were working with Microsoft. Japan’s deputy chief cabinet secretary, Masanao Ozaki, said it affected the government’s IT infrastructure.

In Europe, some UK police departments asked the public to report crimes by phone instead of email or online, and pharmacies reportedly had trouble reaching GP surgeries about prescriptions, according to The Guardian as cited by CNA. Euronext said its listing teams had email problems, and Maersk warned customers of delayed emails and updates. New Zealand paused visa processing, Immigration New Zealand said. ABC reported that stores lost email-based sales and patients lost access to telehealth services.

What Microsoft says caused it

Microsoft said the extra load on its systems did not come from a rise in traffic. Instead, many servers were unexpectedly unavailable, which pushed work onto the servers that remained. According to the AP, some servers failed automatic health checks and were pulled from service. Microsoft switched to a different health-check method and worked to bring the affected servers back.

When restored servers began to fail again, Microsoft said a process responsible for establishing secure connections had become overwhelmed and stopped handling requests. It said it was preparing a fix and, if that worked, would slowly add traffic back. It had also halted all deployments, redirected traffic away from affected infrastructure and restarted servers. The company said it was still working to find the underlying issue and why changes were not syncing properly.

Australia’s Home Affairs minister, Tony Burke, said Microsoft told him it believed this was similar to the 2024 CrowdStrike failure, an update that went wrong, and not a cyberattack. He said a cyber agreement with Microsoft had been activated so Australia gets priority in restoration.

Status of the recovery

Just after 9 a.m. AEST on Nov. 13, Microsoft reported a start to a “gradual recovery.” About an hour later it said restored servers were becoming unresponsive again. Its latest update reported “positive results for server stability,” after it had earlier described only marginal success. ABC reported access to Outlook was being restored gradually, including in Australia. No full resolution time has been given.

Downdetector counts differ by report and are user submissions, not a measure of affected users. ABC cited more than 120,000 Outlook reports on Downdetector US at the peak, around 4:20 a.m. AEST. The AP, as relayed by Chosunbiz, put Outlook reports near 130,000 shortly after 1 p.m. Eastern on Nov. 12.

Why it matters beyond email

The disruption also reached AI tools that connect to Microsoft accounts. OpenAI said it was investigating broken new connections in its Microsoft connector. Anthropic said an issue with an upstream provider was affecting its Microsoft 365 connector, and its status page listed an incident on Nov. 12. ChatGPT and Claude themselves stayed up, but features such as searching email or pulling calendar items through Microsoft were affected.

The outage is the second Microsoft 365 disruption this year, CNA noted, after one in January. Microsoft also reported an Azure outage last month. People who cannot reach their mailbox can check Microsoft service health and basic Outlook fixes, while Microsoft continues to post updates on X.

Sources: ABC (Australian Broadcasting Corporation), CNA, Chosunbiz