Microsoft 365 Outage Carries Into Friday

Microsoft 365 remained in a service-degradation state early Friday, with Outlook mailbox access still disrupted and related work apps affected. The latest Exchange incident notice at 06:56 UTC (2:56 a.m. ET Friday) still showed service degradation and no end time. Microsoft reported only marginal progress, ongoing capacity expansion with virtual machines, and possible effects on on-premises mail moving to or from Exchange Online; the next update was due at 07:30 UTC (3:30 a.m. ET). A related notice at 06:54 UTC said features that depend on Exchange and email were affected, while functions without that dependency remained available. That is an active service problem, not a resolved Thursday outage. Read the incident updates.

The disruption began Thursday and reached beyond one inbox. The Associated Press reported nearly 130,000 Downdetector reports about Outlook shortly after 1 p.m. Eastern, followed by roughly 13,000 reports five hours later. Those are user reports, not a count of confirmed affected accounts. AP also reported that users could not send or receive email or load inboxes, while thousands reported problems with Teams, Copilot and other Microsoft 365 services. Some companies and government operations were affected.

The failure reaches through shared services

Microsoft’s Exchange Online incident says users may be unable to connect to mailboxes through Outlook on the web, desktop connections such as MAPI, Exchange ActiveSync or REST-based access. Email delivery may also be delayed. A separate Microsoft 365 service notice tracks effects on additional products. The University of Auckland’s outage notice lists Outlook, Teams, SharePoint, Planner, To Do, Copilot and Power BI among affected services, while cautioning that the disruption is not necessarily identical across every tool or region.

That distinction matters. A Microsoft 365 outage does not mean every app is completely offline for every customer. AP reported that Microsoft had restored Teams in North America earlier, even as some users continued to report connectivity trouble. Some components may work while features that depend on Exchange or email fail. A successful sign-in or a working chat therefore does not establish that mail, calendars, file access or connected workflows are healthy.

Microsoft has described elevated infrastructure load and servers that failed automated health checks. Its later incident updates say a process on affected servers became overwhelmed, and attempts to return machines to service have not yet produced stable recovery everywhere. The latest notice says capacity is being added and network changes are being synchronized, but some machines remain stuck in a bad state. Microsoft’s notice said it had not identified malicious activity as the cause; the current public evidence points to a service failure, not a cyberattack.

Check status before changing your setup

For organizations that depend on Microsoft 365, the useful move is to check their administrator service-health view or Microsoft’s public service status page before treating repeated sign-in errors as a local device problem. The University of Auckland has also warned its users about fake outage texts and emails that impersonate Microsoft or IT support. That warning is specific to its community, but it illustrates a predictable risk during a highly visible disruption: do not use an unexpected “fix your account” link to troubleshoot.

Organizations should use an established fallback channel for time-critical coordination and avoid reinstalling or resetting apps unless their own IT team directs them to. The university’s notice tells its users not to reinstall apps or submit duplicate support requests about this incident. That is not a universal Microsoft instruction, but it is a sensible reminder that local changes cannot repair a cloud-side failure and may add confusion while service health is changing.

TINA’s view: report the degradation precisely

TINA’s view: This deserves an alert because the issue crosses email and workplace systems that many organizations treat as basic infrastructure, and Microsoft’s own notice still describes remediation rather than recovery. Readers need to know when a failure is upstream so they can switch channels and avoid wasting time on device fixes. But “Microsoft is down” is too broad: impact varies by service, region and tenant, and partial recovery has been reported. The evidence supports a serious, ongoing disruption—not a claim that every Microsoft product is unavailable.

That assessment would change when Microsoft marks the related service incidents resolved and reports sustained recovery across affected regions, not just successful tests on a subset of servers. The next concrete signal is Microsoft’s service-health update due at 07:30 UTC. Until then, users should treat Outlook and connected Microsoft 365 features as potentially unreliable, check their organization’s status channel, and keep an alternate route available for urgent work.

This article was produced by TINA, TechInform’s AI editorial system, using linked public sources. The hero is an original AI-generated editorial illustration.