Confirmed outage — cause still unclear

Microsoft 365's Global Outage: What We Know (and What's Still Unconfirmed)

On November 12, 2026, Exchange Online, Outlook, Microsoft Teams, Copilot and the Microsoft 365 admin center became unreachable for large numbers of users across more than 70 countries. Microsoft says it wasn't a cyberattack or a malfunctioning AI agent — but the official incident report detailing the exact cause still hadn't been published at the time of writing.

November 14, 2026 · SunuCyberSecurity Team

Partager :WhatsAppFacebookLinkedInX
70+

countries that reported Microsoft 365 disruptions

24h+

of disruption before a return to normal was announced on November 14 (some outlets report well over 24h)

2

incident IDs opened by Microsoft: EX1550543 (Exchange Online) and MO1550821 (broader Microsoft 365)

0

evidence of a cyberattack or a malfunctioning AI agent, per Microsoft

Sources: SecurityWeek, CNBC, Time, MacRumors, CBS News, CIO.com — full list at the end of this article

Microsoft first acknowledged a problem at 17:54 UTC on November 12, affecting access to Exchange Online mailboxes for users in the Americas. Thirteen minutes later, at 18:07 UTC, the company widened its statement to a broader issue affecting Microsoft 365 as a whole — Outlook, Teams, Copilot, the admin center, Microsoft Intune, and, per some technical reporting, Microsoft Defender for Office 365 and SharePoint Online. What follows synthesizes what specialized outlets were able to verify and what Microsoft itself publicly stated, keeping the two explicitly separate.

Timeline: two incident IDs, one outage that kept stretching on

Two separate tickets circulated in Microsoft's communications: EX1550543 for the Exchange Online portion, opened around 11:17 AM ET (16:17 UTC), and MO1550821 for the broader Microsoft 365 incident, opened at 1:26 PM ET (18:26 UTC) — a roughly two-hour gap between the two that lines up with the scope widening, rather than two separate outages.

Microsoft then announced it was returning some affected servers to service on the evening of November 12, followed by "partial recovery" in some regions on the morning of November 13, then mail queues "approaching normal levels" later that day. Full restoration was announced Saturday morning, November 14, with no exact time given. On total duration, sources diverge slightly: Time, CNBC and MacRumors describe "almost 24 hours" of disruption, while CBS News and Digital Trends describe an outage that ran noticeably longer, with Digital Trends calling it Microsoft's longest outage in years. The gap likely comes down to different start/end reference points across outlets (first regional report vs. global incident, partial vs. full restoration), not a substantive disagreement on the facts.

The cause, per Microsoft — and what's still unclear

Microsoft attributes the incident to settings applied to servers after an update. Per SecurityWeek, the company said a process responsible for establishing secure connections became overwhelmed, causing a portion of servers to stop responding — those servers reportedly then failed health checks and were pulled out of rotation, overloading the rest of the fleet in turn. Microsoft began rolling back a recent network configuration change identified as a factor.

Worth stating precisely: the timing match between that network configuration change and the start of the outage is real, but SecurityWeek is explicit that, as of its reporting, Microsoft had neither confirmed a definitive root cause nor found evidence of a cyberattack. Microsoft has said a full incident report (Post-Incident Report) would be published — that document, which normally carries the verified technical detail, had not yet appeared at the time of writing, per CIO.com. What's presented here is therefore Microsoft's own account, not an independent, final technical analysis; we'll update this article if the official report tells a different story.

What we know — and what we still don't

Explicitly confirmed by Microsoft: this wasn't a cyberattack, and no AI agent caused the incident — Time reports the company said it has "no evidence of malicious activity associated with the incident." CNBC reports Microsoft publicly apologized for the outage, and CBS News reports the company announced it had "restored" service and addressed the "underlying issue."

Still unclear: the exact total duration (between "almost 24h" and "well over," depending on the outlet), the precise time of full restoration, the complete technical mechanism behind the failure, and, above all, the absence, at the time of writing, of an official incident report that would settle the partially-overlapping accounts from different sources.

Why this matters for francophone Africa

Microsoft 365 is one of the most widely used email and collaboration tools among businesses, banks and public administrations across francophone Africa. None of the sources we consulted publish a country-level breakdown of the real impact in West or Central Africa — the figures reported (70+ countries affected globally) don't let us isolate the local picture, and we found no region-specific reporting at the time of writing. We're stating that plainly rather than asserting it without proof.

What is directly transferable, regardless of this specific outage's real local scale, is the risk category itself: dependence on a single cloud vendor for email, collaboration and, very often, authentication (Entra ID backs the identity of many other business applications) creates a single point of failure. For institutions subject to BCEAO/COBAC frameworks, a business continuity plan that doesn't explicitly cover the scenario "our critical cloud vendor is unreachable for more than a day" leaves a real blind spot — see our article on what an automated scan can actually verify for BCEAO/COBAC compliance for the full methodology.

What to do about it

Don't route all emergency communication through the same vendor

If Teams and Outlook go down together, how do you reach your teams? A backup channel outside the Microsoft ecosystem (SMS, WhatsApp Business, a dedicated phone line) needs to be defined and tested before the incident, not during it.

Check status on the official portal, never via a link received by email

Major outages are reliably followed by a spike in phishing exploiting the confusion ("your account needs verification," fake "Microsoft support" emails). Check status.office.com or the admin center directly, never a link received by email about it.

Build cloud-vendor dependency into your business continuity plan

A BCP that only covers a datacenter fire or a power outage has a real blind spot: "our email and cloud identity are unreachable for more than a day" is its own scenario — especially for institutions subject to BCEAO/COBAC frameworks.

Map what depends on the same cloud identity

When Entra ID, the identity behind Microsoft 365, is disrupted, every application that authenticates against it via SSO is disrupted too — often tools that look, at first glance, unrelated to Microsoft. Knowing which ones before the incident beats discovering them during it.

Sources and methodology

SecurityWeek (technical detail on the overwhelmed secure-connection process, no confirmed root cause) · CNBC (Microsoft's public apology) · Time (no evidence of malicious activity) · MacRumors (timeline, ruling out an AI agent) · CBS News (restoration announcement) · CIO.com (no official incident report published yet). Outage observed between November 12 and 14, 2026. No official Microsoft Post-Incident Report was available at the time this article was published — we'll update it if Microsoft releases further detail.

Related reading

BGP Hijacking: How an ISP Can Protect Its IP Prefixes What an automated scan can actually verify for BCEAO/COBAC compliance (and what it can't)

Check your own exposure

A free scan, in minutes, no credit card required.

Free scan

Newsletter

Get our next analyses

One email whenever we publish a new article. No spam, one-click unsubscribe.