Microsoft 365 went down worldwide: the single-vendor lesson for your business
Direct answer
On November 12, 2026, Microsoft 365 suffered a worldwide outage that started in Exchange Online, the email service behind Outlook for most organizations. Mailboxes stopped opening at around 16:17 UTC, and the fault spread to Outlook, Teams, SharePoint, Copilot and Microsoft Defender, among other services. Downdetector logged more than 120,000 Outlook problem reports at the peak, from 74 countries, and the outage was Microsoft's largest in 12 months according to the platform's data. Microsoft declared the service restored in the early hours of Saturday, November 14, about 24 hours after it began. The company traced the fault to a problem that surfaced during a network capacity upgrade and ruled out a cyberattack.
What went down, and in what order
- Email went down first: users stopped opening their Exchange Online mailboxes. Microsoft opened incident EX1550543 for Exchange and the general incident MO1550821 for Microsoft 365.
- The fault spread through dependency: because so many Microsoft 365 services depend on Exchange Online, the impact reached Teams calendars and scheduling, SharePoint features, Microsoft 365 Copilot, Microsoft Graph, Planner, To Do, Intune, Purview and Power BI.
- Teams calls mostly kept working: joining calls, meetings and chats in Teams worked in general. What failed was what depends on email and calendars: scheduling, inviting and editing meetings.
- The AI lost its email too: OpenAI reported trouble connecting ChatGPT to Microsoft 365, and Anthropic said mail, calendar and search requests in its Microsoft 365 connector and Claude for Outlook returned errors for more than 24 hours.
The cause, according to Microsoft
- A network capacity problem: Microsoft said a problem during an earlier network capacity upgrade surfaced after a recent network change. The company began reverting that recent change, which lined up with the start of the fault.
- Cyberattack ruled out: Microsoft executive Charles Lamanna ruled out a cyberattack, and Microsoft said it found no evidence of malicious activity.
- The full report is still pending: Microsoft posted its final update confirming full recovery and processing of the delayed email backlog, and announced a post-incident report with the complete review. At the time of writing, that detailed report was still pending.
- Delivery timeout extended to 48 hours: Microsoft extended the email delivery timeout to 48 hours so delayed messages would not bounce while the backlog was processed.
What to do in your business this week
- Map your dependence on a single vendor: write down which parts of your operation depend on Microsoft 365: email, calendar, files, meetings and your Copilot. The outage showed that a fault in email also stops the calendar and the AI that reads that email.
- Define a backup channel for email: a second communication channel for the team and for customers, with key contacts stored outside Outlook. The backup gets tested before a crisis, with a short send-and-receive exercise.
- Write the plan for the next outage: who tells customers, through which channel, and which work continues without email. A plan written on one page beats improvising with email down.
- Review the permissions of your connected agents: if your AI agents read email and calendars through connectors, define what they do when the connector returns errors: retry with a limit, log the failure and tell a person. An agent that retries without a limit during an outage creates noise and burns quota.
Note: this article summarizes Microsoft 365's public status updates (incidents MO1550821 and EX1550543), Downdetector data, and coverage by BusinessDay, Wire, Acorn Technology, KeyXplore and 4MIPS on November 13 and 14, 2026. Report figures come from Downdetector. The cause is Microsoft's published explanation; the detailed post-incident report is still pending.
Want AI that works for your business?
Message us on WhatsApp and let's talk about automation and AI agents for your business.
WhatsApp inicIARelated articles
- Google launched the universal Gemini agent: the digital coworker with its own email
- Claude becomes your workspace: Whiteboards and Sheets arrive in beta
- Anthropic opens the beta of self-hosted Claude Managed Agents: the agent runs in your cloud
Frequently asked questions
What went down in Microsoft 365 on November 12, 2026?
The fault started in Exchange Online, the email service behind Outlook, and spread to Outlook, Teams, SharePoint, Microsoft 365 Copilot, Defender and other services that depend on email and calendars. Teams calls mostly kept working.
Was it a cyberattack?
Microsoft said it found no evidence of malicious activity, and executive Charles Lamanna ruled out a cyberattack. The company traced the fault to a problem that surfaced during a network capacity upgrade.
How long did the outage last?
Mailboxes began failing on Thursday, November 12, at around 16:17 UTC, and Microsoft declared the service restored in the early hours of Saturday, November 14, about 24 hours later, with the delayed email backlog processed.
What do I do if my business depends on Microsoft 365?
Map which parts of your operation depend on that vendor, define a backup channel for email and customers, write a one-page plan for the next outage, and set retry limits on AI agents that read email and calendars through connectors.
