Neowin, citing Microsoft's announcement, reported that users in Germany and the United Kingdom are currently experiencing connectivity issues when using Microsoft 365 services due to degraded Global Secure Access (GSA) infrastructure performance. The reported list of affected services covers most of a normal workday: Outlook, Outlook on the web, OneDrive, SharePoint Online, Microsoft Teams, and Windows 365 Cloud PCs.
What Microsoft has said so far
Here is the status as of Neowin's report, published at 12:51 UTC on October 13:
- Cause: Microsoft says it has found the section of GSA infrastructure behind the slowdown. It is applying performance and traffic optimizations to "remediate impact and restore connectivity."
- Start time: The problem had been running since at least 11:43 AM UTC. At publication, that comes to about 68 minutes of reported impact.
- Next update: Microsoft said its next update would come at 1:30 PM UTC.
- Resolution: None confirmed yet. No end time and no confirmation of full recovery were available when this was written.
Microsoft's public statements cover Germany and the UK only. Community outage trackers suggest people elsewhere noticed problems too. User reports on StatusGator's Azure Global Secure Access page, dated October 13, include "GSA down" from North Rhine-Westphalia, a Zurich user writing "GSA Down in Germany," and a Danish user posting "GSA dead, Denmark." These are unverified reports from users. They do not show that Microsoft's official impact area extends past the two countries it named, so it's too early to call this a Europe-wide outage.
Section summary: Microsoft has acknowledged a GSA-related Microsoft 365 connectivity incident affecting the UK and Germany and is rolling out mitigations. It had not declared recovery when this was written.
Why a GSA problem knocks out Outlook and Teams
Global Secure Access is Microsoft's Security Service Edge offering, part of the Microsoft Entra family. Its Microsoft traffic profile does what the name suggests. In Microsoft's own operations documentation, the profile routes Microsoft 365 and other Microsoft service traffic through GSA infrastructure, where security controls and routing optimizations are applied.
So for organizations that have turned that profile on, GSA sits directly in the path between a user's PC and Exchange Online, SharePoint, Teams and related services. If that path slows down, the apps slow down or stop with it, even when the Microsoft 365 services behind it are healthy.
Microsoft's guidance already warns about this. Its GSA operations page says the Microsoft traffic profile is unusual because it carries traffic for services organizations use every day (Teams, Exchange, SharePoint), so any disruption has immediate, visible user impact. The same document tells admins to treat a disabled Microsoft traffic profile as a critical incident, because Microsoft 365 traffic would then stop going through GSA.
Our analysis, drawn from general industry experience rather than anything Microsoft has said about this incident: this is the standard trade-off with cloud security proxies. Sending all your traffic through one inspection point gives you consistent policy, compliant-network signals for Conditional Access, and tenant restrictions. It also gives you one more dependency that can fail. Today's incident is that dependency failing.
Section summary: GSA can sit in front of Microsoft 365 traffic, which is how a network-layer problem shows up as broken Outlook, Teams and Cloud PC sessions.
Not the first UK-flavored Microsoft 365 wobble this autumn
This is the second geographically scoped Microsoft 365 incident hitting UK users in under a month. In mid-September, WindowsForum covered single sign-on failures in which Microsoft 365 users in the United States and United Kingdom saw sign-in failures and 5xx server errors. That incident was tracked as MO1472904 in Microsoft's service-health communications, and earlier status updates described the scope as U.S.-only before the title was expanded to include the United Kingdom.
The UK–Germany pairing has also happened before. In an older post on its Microsoft 365 Status account, Microsoft said it was investigating an issue where some users in the UK and Germany may be unable to access Microsoft 365 services.
The September case is a useful warning for today. The first geographic description of an incident isn't always the final one. Admins outside the UK and Germany who see similar symptoms should check their own tenant's Service health page instead of assuming they're unaffected.
What admins should do right now
None of the steps below is a Microsoft-approved fix for this specific incident. They come from Microsoft's documented procedures for checking service health and running GSA.
1. Check Service health in the admin center
- Sign in to the Microsoft 365 admin center with an admin account.
- Go to Health > Service health, or select the Service health card on the Home dashboard.
- Look under Active issues Microsoft is working on and open the issue to see its full update feed.
The issue detail page shows the incident ID, estimated start time, affected services, issue type, issue origin, status and user impact. Accounts with the Service Support admin or Helpdesk admin roles can view service health. If you can't get into the admin center at all, Microsoft points admins to its public service status page.
2. Report impact Microsoft hasn't listed
If your users are affected but your tenant doesn't show a matching issue, select Report an issue and fill in the form. Microsoft says it compares those reports with data from other organizations to judge how widespread a problem is and whether it started on Microsoft's side.
3. Subscribe to updates
- To follow just this incident, open the active issue and select Manage notifications for this issue.
- For future incidents, go to Customize > Email, turn on service health notifications, and pick the services and issue types you care about. Each admin account can add up to two email addresses.
4. Check whether GSA is actually the factor
Microsoft's GSA operations guidance tells help desks to compare user-reported Microsoft 365 performance problems with GSA traffic logs, to see whether GSA routing is involved. It also recommends correlating GSA telemetry with Microsoft Entra sign-in logs, and using the Global Secure Access workbook in Microsoft Sentinel for traffic and denied-session trends.
If some of your sites use GSA and others don't, comparing them can tell you quickly whether you're dealing with this incident or a separate problem.
5. Think before you turn GSA off
When the security proxy is the thing breaking, switching it off is tempting. Be careful. Microsoft hasn't published that as a workaround for this incident. If you rely on GSA compliant-network signals in Conditional Access, bypassing GSA could start blocking sign-ins that depend on those signals. It could also leave Universal Tenant Restrictions unenforced, which Microsoft's documentation lists as a critical condition in its own right. Any routing change should go through your change process and a security review.
Section summary: Use Service health and the per-issue notifications to follow the incident, check GSA logs against user reports, and don't treat bypassing GSA as a quick fix.
The bigger picture
For organizations in the affected areas, the practical cost is easy to picture. The incident started around midday local time, so many people lost Teams calls, shared files, mail and Cloud PCs for a good part of their afternoon. The full business impact, and whether this lasts hours or is fixed quickly, depends on how well Microsoft's traffic optimizations work. At the time of writing, that's unknown.
There's also a broader question. Microsoft has been pushing its Security Service Edge stack as the modern, zero-trust way to reach Microsoft 365. Incidents like this one don't undercut that approach, but they show the cost: the more traffic you send through a single cloud inspection layer, the more an outage in that layer can take down. Expect admins to be asking about resilience and fallback routing at the next architecture review.
WindowsForum will update this story as Microsoft posts further status updates, including the one it scheduled for 1:30 PM UTC.