Skip to the cards
Filed
Source
Suhasini Srinivasaragavan
Loans
414

Microsoft 365 disruption spans Outlook, Teams, Exchange and sign-in services

Microsoft reported elevated load after servers became unavailable; user reports show disruption across Microsoft 365 services in 74 countries.

Microsoft 365 disruption spans Outlook, Teams, Exchange and sign-in services

A Microsoft 365 disruption lasting more than 12 hours had affected hundreds of thousands of users worldwide by 13 November 2026, with reports spanning email, collaboration, spreadsheets and account access. For administrators, the breadth matters: Downdetector users reported problems with Exchange, Excel and sign-in functions alongside Outlook, Teams, Copilot, Defender and education and admin services.

A smartphone screen with Microsoft Office apps.

Microsoft attributed the incident to strain across parts of its infrastructure. In a post on X on 12 November, the company said several servers had become “unexpectedly unavailable,” leaving the remaining infrastructure under elevated load across multiple regions. It said it had seen “marginal success” in its response at the time of its service health update and was continuing to monitor server status across regions.

Disruption reaches core work and access services

Reports tracked by Downdetector indicated that Exchange, Excel and login functionality were among the most affected areas. The disruption was reported in 74 countries and included services used for email, meetings, productivity work, security and tenant administration. The available information does not identify specific Microsoft 365 licence tiers, customer configurations or affected service components, so administrators should check their own tenant’s service health details before drawing conclusions about local impact.

Revolut services affected in fresh Google Cloud outage

In Ireland, user reports rose sharply at around 6pm on 12 November. Approximately 500 reports from Ireland remained visible at the time of publication. That figure represents reports submitted to Downdetector; it is not a count of verified affected users or a measure of the total number of customers experiencing disruption. The broader estimate of hundreds of thousands of users also reflects outage reporting, rather than a Microsoft count of impacted accounts.

Anthropic challenges Microsoft with new Claude office tools

The incident’s effects may differ between organizations. Some users may be unable to sign in or reach mail, while others could encounter problems in Excel, Teams, Copilot, Defender or administrative tools. The report does not specify whether these symptoms were continuous, how they varied by region, or whether they affected desktop, browser and mobile clients equally. That leaves tenant-level service health as the more useful guide for determining which workloads are impaired in a particular environment.

Cloud distributor Sherweb picks Dublin for European HQ, plans jobs

Microsoft says it has no indication of malicious activity

Microsoft said it was not aware of malicious activity causing the incident. Its description instead pointed to unavailable servers and rising load on the infrastructure still operating. That distinction is useful for incident triage: the company’s stated working explanation was a service availability problem, but it had not presented a detailed technical cause or a confirmed recovery timeline in the updates cited here.

Household names co-sign open letter on AI cybersecurity threat

For IT teams, Microsoft’s “marginal success” wording signals that recovery work was under way but that progress remained limited at that point. The company said it would continue monitoring activity across regions and assessing server status. No specific restoration deadline, workaround, command or configuration change was provided in those updates, so administrators should avoid treating client-side changes as a confirmed remedy.

Reuters: China’s Moonshot talking to US cloud giants over Kimi K3

A practical response is to establish which users and workloads are affected, then compare those reports with the tenant’s service health information. Help desks can record the service, region, approximate start time and error or symptom reported by each user. That gives administrators a clearer view of whether an issue is concentrated in sign-in, Exchange, Teams or another named workload, and helps them update staff without implying that every Microsoft 365 service is unavailable.

Report: Only sectors aiding AI adoption sure to grow from it

Where multiple services fail together, administrators may also need to account for dependencies in their own work processes. A sign-in problem, for example, can prevent access to several services even if each workload is not independently impaired. The incident information does not confirm that this was the cause of any particular user’s symptoms; it is a reason to distinguish access failures from service-specific faults when collecting reports.

Third-party services can feel the impact

The disruption extended beyond Microsoft’s own products. OpenAI said it was investigating an issue with Microsoft connectors, showing how a Microsoft service problem can affect integrations that rely on the company’s infrastructure. Teams that depend on connectors or other linked services may therefore see secondary symptoms even when those services are operated by another provider.

Recent cloud disruptions have also affected businesses that depend on shared infrastructure. A Google Cloud incident in western Europe in October 2026 was linked to service effects at Revolut, Lloyds Bank, HSBC, Royal Mail and Cloudflare. Major outages involving Cloudflare, Microsoft Azure and Amazon Web Services were also reported the previous year. These examples underline why incident teams need to check both the primary service and connected providers when users report failures.

For now, the concrete checks are to review Microsoft 365 service health, identify which users and workloads remain affected, and track Microsoft’s regional updates for evidence of recovery. The company had not identified malicious activity as the cause in the information available on 13 November, and its response remained focused on server availability and elevated infrastructure load.

Source: Suhasini Srinivasaragavan · www.siliconrepublic.com

Filed alongside