Skip to the stories
Microsoft 365 & SharePoint

Microsoft 365 outage hits Outlook, Teams and Copilot in 74 countries

A Microsoft 365 disruption lasting over 12 hours hit Outlook, Teams, Copilot and Defender across 74 countries. Here's what admins should check now.

Microsoft 365 outage hits Outlook, Teams and Copilot in 74 countries

Microsoft 365 has been degraded for more than 12 hours, and the scale is what makes this one worth a close look. Downdetector recorded reports from hundreds of thousands of users, with Exchange, Excel and login functions the most affected and disruption reported in 74 countries. For anyone running a tenant, the practical question is less about the headline and more about which of your workloads depended on the parts of the service that failed.

A smartphone screen with Microsoft Office apps.

What broke and where

The disruption reached well beyond email. Outlook, Copilot and Teams were affected, along with Microsoft 365 Education, admin functions and the Defender security tools. Downdetector's data put Exchange, Excel and sign-in as the most affected areas. That mix matters. A mail outage is painful but usually obvious. A sign-in problem or an admin-portal problem can stall the people who would normally diagnose and communicate the issue, which is why many teams found the incident harder to manage than a single-app failure.

Reports from Ireland spiked at around 6pm on 12 November, and at the time of publication the count there sat at roughly 500. Those figures come from user-submitted reports on Downdetector, so they show how many people noticed a problem rather than the number of affected accounts. Treat them as a signal of breadth, not a measure of impact.

Revolut services affected in fresh Google Cloud outage

What Microsoft has said

Microsoft posted on X on 12 November that a portion of its infrastructure was under elevated load across multiple regions. The company said a number of servers had become "unexpectedly unavailable", which pushed extra strain onto the infrastructure still running. Its service health update added that Microsoft was not aware of any malicious activity behind the incident, and that it was watching activity across regions to evaluate server status. In the same update it described having "observed marginal success", which reads as progress without a recovery that admins can count on yet.

Anthropic challenges Microsoft with new Claude office tools

Microsoft has said it does not suspect foul play at this time. For admins, that points toward an availability and capacity fault rather than a security event. It doesn't mean you can skip the security checks, though. Sign-in anomalies and Defender alerts during a degraded period deserve the same scrutiny you'd give them on a normal day.

Cloud distributor Sherweb picks Dublin for European HQ, plans jobs

Knock-on effects for connected services

The impact spread into third-party services that rely on Microsoft's connectors. OpenAI said it was investigating an issue with Microsoft connectors during the same window. If your organisation has built automations, integrations or AI workflows that call Microsoft 365 APIs, expect failures to look like application errors on your side. Retries, queued jobs and webhook deliveries can pile up and then fire in bursts once service returns, which can create a second wave of problems.

Household names co-sign open letter on AI cybersecurity threat

Mail-dependent systems were also affected. A New Zealand ministry reported that its ability to receive email had been hit, and several public-sector bodies in other countries were affected too. If you run line-of-business applications that send notifications through Exchange Online, or that depend on inbound email for approvals and ticketing, check those flows specifically rather than assuming they recovered with the rest of the service.

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

What to check in your own tenant

Start with the basics. Check the Microsoft 365 service health dashboard and the message centre for your tenant, because the public status page and your tenant's view of its own services don't always match. Note the start time and the services your users reported, so you have a clean timeline for your incident record and for any service-credit claims your contract allows.

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

Next, look at what's queued. Exchange mail that was deferred or delayed may still be in transit, and integrations may have failed silently. Confirm that critical flows have actually completed rather than trusting that a green dashboard means the backlog has cleared. Where your team has a fallback channel for urgent contact, make sure people know how to reach each other if Teams and email are both down.

Finally, review how your admin tooling behaved. Some incidents hit the same portals you'd use to respond. If your team couldn't open the admin centre or sign in cleanly, that's a dependency worth documenting and planning around, even if the outage itself is resolved.

Repeat incidents, and what to watch next

This wasn't an isolated event for the cloud sector. Last month a disruption to Google Cloud's services in western Europe affected a range of businesses, including Revolut, Lloyds Bank, HSBC, Royal Mail and Cloudflare. Cloudflare, Microsoft Azure and Amazon Web Services also all faced major outages in the previous year. That pattern is the real lesson. A single provider's bad day can take down services you assumed were independent, because so many of them sit on the same few platforms.

My read is that the useful response is practical rather than dramatic. Map which business processes depend on Microsoft 365 sign-in and Exchange, decide what your team does when they fail, and make sure those steps are written down and tested. Watch for Microsoft's final incident report, which should give the root cause and the scope of the server-availability problem. That report will tell you more about whether this was a one-off capacity event or something that could recur, and it's the document you'll want before the next change window.

Source: Suhasini Srinivasaragavan · www.siliconrepublic.com

Same flag