Microsoft Faces 12-Hour Outage Affecting AI Services

Microsoft 365, Outlook, and Teams experienced access problems in several regions after parts of Microsoft’s infrastructure became overloaded. Microsoft said some servers failed automated health checks and went offline, shifting additional demand onto the servers that remained available. The company said the incident was not caused by a surge in user traffic.

Reports on Downdetector began at about 20:30 local time. More than 12 hours later, some users were still reporting problems. The disruption affected users across 74 countries, with broader reports of outages in parts of Asia.

What caused the Microsoft 365 outage?

According to Microsoft’s initial findings, some servers did not pass automated health checks and were taken out of service. This reduced available capacity and increased the load on other parts of the infrastructure. Microsoft also identified a configuration issue affecting the health-check process and tested a change intended to address it.

The reported cause was therefore an infrastructure and server-health issue, rather than unusually high customer demand. Microsoft had not reported widespread data loss in the information available during the incident; the main reported effects involved access and service availability.

Which services and users were affected?

Users reported problems accessing Microsoft 365 services, including Outlook and Teams. The disruption also affected some third-party tools and integrations that rely on Microsoft services or connections to Microsoft 365. OpenAI reported that some tools were unavailable, while Anthropic reported problems with features connected to Microsoft 365.

• Outlook: Some users experienced access problems, delayed email, or unsuccessful delivery attempts.
• Teams: Users reported trouble connecting to meetings and using features such as audio or screen sharing.
• Connected applications: Some third-party services and API-based workflows that depend on Microsoft infrastructure also experienced interruptions.

The exact impact varied by region, service, and organization. An outage in a cloud platform can affect connected tools even when those tools are operated by a different company.

How did Microsoft respond to the outage?

Microsoft said it paused deployment activity, worked to reduce pressure on affected infrastructure, and began restoring servers. Its response included safely restarting servers that could not provide service, testing a configuration change, and applying an alternative health-check method that brought some servers back online.

• Step 1: Pause planned deployments to avoid introducing further changes while investigating the incident.
• Step 2: Reduce the impact of affected infrastructure on other parts of the service.
• Step 3: Restart unavailable servers safely and review their startup status.
• Step 4: Test a corrected configuration and an alternative health-check method.
• Step 5: Restore servers progressively and monitor service stability.

Microsoft’s updates described recovery work in progress. The reports available at the time did not confirm that every affected user had regained full access.

Why can a server health-check issue cause a widespread outage?

Health checks help cloud systems decide whether a server is operating properly. When a server fails a check, the system may remove it from service to protect users from unreliable performance. If many servers become unavailable or fail checks because of a configuration problem, the remaining servers may have to handle more work than usual. That can lead to slower responses or connection failures across multiple services.

Cloud services also share infrastructure and dependencies. As a result, an issue in one part of the platform can affect products such as email and online meetings, along with third-party applications that depend on Microsoft 365 connections.

What did the outage mean for businesses?

Organizations that rely on Microsoft 365 for daily operations could face delayed email, disrupted meetings, and stalled workflows. For example, a team might be unable to join a Teams call while an automated system waits for an Outlook or Microsoft 365 connection to complete. The outage could therefore affect both direct users and processes running in the background.

Businesses can reduce disruption by identifying which essential workflows depend on Microsoft services and deciding how staff should continue work when those services are unavailable.

How can organizations prepare for a cloud-service outage?

• Set up backup communication channels: Keep an alternative method available for urgent messages and critical meetings.
• Map service dependencies: Record which business processes rely on Microsoft 365, Outlook, Teams, APIs, and connected applications.
• Define workarounds: Give teams practical instructions for handling delayed email, unavailable meetings, and interrupted integrations.
• Test continuity plans: Run exercises that simulate a cloud-service outage and check whether staff can carry out essential tasks.
• Monitor business impact: Track failed API calls, email delivery delays, and affected workflows rather than relying only on individual user reports.

How can users check whether Microsoft 365 is working again?

Check Microsoft’s service-status updates and compare them with reports from affected users. Organizations can also review internal error logs, API failure rates, email queues, and delivery delays. A service may appear to be recovering while particular regions, accounts, or integrations continue to experience problems.

Frequently asked questions

How long did the Microsoft 365 outage last?
Initial reports appeared at about 20:30 local time. Some users continued to report access problems more than 12 hours later. The duration varied by user and service.

Was the outage caused by heavy traffic?
Microsoft said the incident was not caused by increased traffic. Its initial investigation pointed to servers failing automated health checks and the resulting increase in load on other infrastructure.

Were Outlook and Teams affected?
Yes. Users reported issues with Outlook and Teams, including access problems, delayed email, and difficulties connecting to meetings or using some meeting features.

Did the outage affect OpenAI or Anthropic tools?
Some tools and features connected to Microsoft services were also reported as unavailable. The effect depends on each product’s reliance on Microsoft infrastructure or Microsoft 365 connections.

Was user data lost?
The incident information described access and availability problems. Microsoft had not reported widespread data loss in the updates summarized here.

Key takeaway

Microsoft attributed the disruption to server health-check failures and a configuration issue, which reduced available capacity and affected connected services. Microsoft paused deployments, restarted servers, tested configuration changes, and applied an alternative health-check method. For businesses, the incident highlights the value of backup communication channels, documented service dependencies, and tested continuity procedures.