Microsoft cloud outage takes Outlook, Teams, and SharePoint down for hours

What happened, hour by hour

On Thursday, November 12, 2026, Microsoft 365 suffered an outage that knocked out business customers' access to several of the suite's services at the same time. The first public confirmation came around 10:40 a.m. (New York time, or 12:40 p.m. Brasília time), when Microsoft posted on X that it was "investigating reports of issues accessing Microsoft Teams in North America".

Shortly before 1 p.m. ET (3 p.m. Brasília time), the company expanded the alert: the issue also affected access to Exchange Online inboxes "through multiple connection methods across the Americas." The peak in Downdetector reports happened around noon ET (2 p.m. Brasília time) and began to drop around 1:15 p.m. ET (3:15 p.m. Brasília time), according to the monitoring site cited by Wired.

About an hour later, Microsoft said it had begun redistributing connection traffic across its global infrastructure to balance the load. The company also published an update explaining the actions underway:

We've paused all deployments. We're moving additional traffic off the affected infrastructure, and we're restarting affected servers to return them to service.

Microsoft, in an update on the Microsoft 365 status page

Around 3 p.m. ET (5 p.m. Brasília time), Microsoft detailed the technical cause visible up to that point: "an issue is causing servers to fail automated health checks and become unavailable," and the company was trying to restart the systems to bring them back online.

Which services were affected

Microsoft's status page listed the following as potentially impacted:

• Exchange Online (corporate email via Outlook)
• Microsoft Teams
• SharePoint Online
• Microsoft Copilot 365
• Microsoft Defender XDR

Microsoft also said its network was "experiencing elevated load in multiple regions, which likely contributed to the connectivity errors," but was careful to separate the two issues: the root cause was servers going offline, not a traffic spike from excessive access. In a statement to Wired, the company reinforced the point:

We are aware and investigating reports of users experiencing issues with Microsoft 365 services, primarily issues with Exchange Online. We appreciate our customer's patience while we work to restore services as soon as possible.

Microsoft spokesperson, in a statement to Wired

What remains unclear

As of Wired's report deadline, Microsoft had not publicly confirmed the root cause of the outage or given an estimated time for full resolution. The report notes that the company did say it was "aggressively pursuing every avenue to reduce the impact," which indicates the investigation was still underway at the time of publication.

Wired notes that cloud service outages are common and have varied causes: configuration errors, DNS resolution problems, and cyberattacks are among the most common possibilities. In this Thursday's case, the report is explicit: there is, so far, no indication that a cyberattack caused the problem.

Not the first time: the July 2024 precedent

Large-scale Microsoft outages are uncommon, but far from unprecedented. In July 2024, a corrupted CrowdStrike update caused the "blue screen of death" on Windows machines worldwide. Almost at the same time, a separate outage took down Microsoft Azure services for hours, generating losses that Wired describes as "countless millions of dollars" in revenue for companies that were unable to operate normally. Wired was careful to clarify that this Azure episode was not directly related to the CrowdStrike problem on Windows, despite the coincidence of dates.

The parallel matters because it shows a pattern: when Microsoft 365 or Azure go down, the cascading effect simultaneously hits email, chat, shared documents, and, increasingly, the AI tools built into the suite, such as Copilot 365. It is not an isolated failure of a single app; it is the failure of a platform that entire teams depend on to operationally exist.

What changes for those running infrastructure in Brazil

The outage was reported as concentrated in the Americas, which includes Brazilian operations that run Exchange Online, Teams, or SharePoint as the backbone of communication and collaboration. For engineering and DevOps teams here, the episode reinforces concrete points for continuity planning:

• Single-vendor dependency for critical communication services. If the company's incident runbook depends on Teams to coordinate incident response, and Teams itself is down, the team loses its communication channel at the worst possible moment. It's worth having a documented and tested fallback channel (Slack, Signal, even a WhatsApp group agreed on in advance).
• Status pages as a source of truth, not a late warning. Microsoft updated the Microsoft 365 status page and X several times throughout the day. Monitoring these sources programmatically (via webhook or RSS, when available) is more reliable than waiting for the internal help desk to identify the problem.
• SLAs and contracts don't eliminate operational risk. Even enterprise customers with a locked-in SLA experience the blackout like any other user for as long as the incident lasts; the contract guarantees financial credit afterward, not service continuity during the outage.
• Authentication and SSO tied to the same provider. If the organization's login runs through Azure AD / Entra ID and other systems (VPN, internal tools) depend on that same SSO, a Microsoft 365 outage can lock access to systems that, technically, aren't even Microsoft's.

The July 2024 case had already raised this warning about risk concentration in a few cloud providers; this Thursday's outage is a reminder that the problem hasn't been solved, it has just gotten less headline attention until the next time it happens.