- Status: Under investigation — Microsoft says it is actively working to resolve the issue.
- When it started: 1:22 PM UTC on September 16
- What’s affected: Microsoft 365 services (specific products not yet itemized by Microsoft).
- Error code: 5xx server-side errors.
- Reported root cause: Upstream dependencies, which Microsoft is still examining.
Microsoft is dealing with a live service disruption that has left some users unable to reach their Microsoft 365 tools, and the company says the trouble traces back to third-party dependencies it relies on.
What we know about the Microsoft 365 outage
According to Neowin, which tracked the incident, Microsoft opened an investigation after reports of the problem began coming in. The company logged the start of the incident at 1:22 PM UTC on September 16, and its engineering teams are now working to restore the affected services.
So far, Microsoft has pointed to two clues in its investigation: upstream dependencies and 5xx errors. It hasn’t yet confirmed exactly which services are down or pinned down the specific third-party component that appears to be at fault, so details remain thin as the incident develops.
Decoding the 5xx errors and “upstream dependencies”
For readers who aren’t familiar with the terminology Microsoft is using, here’s what it actually means in practice.
5xx errors are a family of HTTP status codes that indicate a problem on the server side rather than with your connection or your browser. When you see a 500, 502, 503, or 504, you’re essentially being told that the service you asked for couldn’t complete the request. It wasn’t a “page not found” situation (that’s a 4xx code) but rather “the server had trouble serving it.”
Upstream dependencies refers to external services, APIs, or infrastructure that Microsoft 365 itself depends on to function. Think of it as a chain: your Outlook or Teams client talks to Microsoft’s servers, which in turn may call on another provider’s service to complete the job. If that downstream link breaks, the whole request fails — even though Microsoft’s own systems are technically running fine.
The practical implication is important: an upstream dependency issue often can’t be fixed by Microsoft alone. It typically requires coordination with the third-party provider, which can extend how long the outage lasts and make it harder to predict when things will fully recover.
What the Microsoft 365 outage means for you
If you rely on Microsoft 365 for work, the impact is straightforward — you may be unable to sign in, send email, access files, or use collaboration tools until the issue clears. The exact experience depends on which services are affected, and Microsoft hasn’t published a full list yet.
The good news is that a dependency-driven outage is usually a “wait it out” situation rather than something you can fix on your end. There’s no client-side setting or restart that will bypass a broken upstream link, so patience is the main tool you have. Repeatedly trying to force a connection rarely helps and can sometimes trigger additional errors.
How to check status and what to do
While the outage is ongoing, here are the most reliable ways to track it and manage the disruption:
- Check the Microsoft 365 Service Health dashboard — admins and users can view real-time status and incident updates through the Microsoft 365 admin center.
- Look at Microsoft’s service status pages — these aggregate health across individual products like Exchange Online, Teams, and OneDrive.
- Watch a third-party tracker — sites like Downdetector surface user-reported spikes, which can help confirm whether an issue is widespread or localized.
- Don’t spam sign-in attempts — repeated log-in attempts during an outage rarely help and can sometimes trigger additional errors.
Context: how common are these outages?
Microsoft 365 is one of the largest cloud productivity platforms in the world, serving hundreds of millions of users and thousands of organizations. Given that scale, brief service interruptions are relatively routine, and most resolve within a few hours without lasting damage.
Dependency-related outages like this one tend to be a reminder of how interconnected modern cloud services have become. Even a well-funded provider like Microsoft can be held hostage by a single third-party link, which is why these incidents often draw more scrutiny than others.
For IT admins managing a fleet, this is a good moment to monitor the Service Health dashboard and, if you route updates through WSUS or Intune, keep an eye on whether the outage is affecting deployment or telemetry traffic.
As of now, Microsoft’s investigation is ongoing, and the company hasn’t provided an ETA for full recovery. We’ll be tracking the situation and updating as more details emerge.
Source: Neowin
Over to you: Will you keep relying on Microsoft 365 for your daily work, or are you starting to look at alternatives after outages like this one?
