Microsoft’s massive outage enters day two, takes ChatGPT and Claude connectors down with it

Microsoft’s significant outage has now reached its second day, disrupting ChatGPT and Claude connectors and prompting worries about the dependability of cloud services.

Related: Enterprise AI startup Ema raises $77 million in Series B funding led by Creaegis

A significant Microsoft 365 outage has disrupted businesses and users in numerous countries, with the effects reaching beyond Microsoft’s own applications. The incident, which started on November 12, 2026, impacted services such as Outlook, Teams, SharePoint Online and Microsoft Copilot.

As the disruption carried on into November 13, third-party AI services were also affected. OpenAI and Anthropic reported issues with the Microsoft 365 connectors that ChatGPT and Claude rely on. These integrations let AI assistants access workplace information like emails, calendars, Teams conversations and documents.

The incident underscored how tightly interconnected today’s digital workplace has become. When a major cloud provider runs into technical problems, applications that depend on its infrastructure may also experience disruptions.

Microsoft reported progress in its recovery efforts, though service availability remained inconsistent. The company kept working to restore affected infrastructure and stabilise its services.

The outage impacted the Microsoft 365 connectors within ChatGPT and Claude. These connectors let users connect their workplace accounts to AI assistants and pull information from business applications.

For instance, users can ask an AI assistant to locate an email, summarize a conversation, check calendar events, or search company documents.

However, these features rely on Microsoft’s services being available. When the underlying infrastructure ran into problems, the connected AI tools were unable to reliably retrieve the information they needed.

Consequences for ChatGPT’s Microsoft 365 Connector

On its service status page, OpenAI posted an incident called Elevated Errors Affecting the Microsoft 365 Connector.

The company cautioned that:

• Attempts to create new connections to Microsoft services may fail.
• Tools that are already connected might not function as expected.
• When accessing connected workplace information, users may encounter errors.

The problem impacted the Connectors and Apps feature of ChatGPT. This did not imply that all ChatGPT features or the underlying AI model were unavailable.

Effect on Claude’s Microsoft 365 Connector

Anthropic additionally reported issues with its Microsoft 365 connector.

The outage impacted multiple workplace functions, among them:

• Access to Outlook email.
• Calendar requests and scheduling.
• Chat searches in Microsoft Teams.
• Document searches in SharePoint.
• The Claude for Outlook integration.

At first, Anthropic pinpointed a problem with an upstream provider. It subsequently verified that the Microsoft 365 service incident was the source of the disruption.

The company also stated that most requests involving its mail, calendar and search tools were failing throughout the incident.

Notably, the disruption did not impact every Claude service in the same manner. Claude API, Claude Code and Claude Console were shown as operational in the status updates that were reported.

The disruption spread across multiple Microsoft 365 products. Nevertheless, the impact differed based on the service and how heavily it relied on the affected infrastructure.

The outage caused especially significant disruption for organizations that depend on Microsoft 365 for everyday communication and collaboration.

Why were some Teams features still functioning?

Not all Microsoft Teams functions ceased to work.

For instance, some users were still able to join existing meetings even while facing issues with calendars or meeting scheduling. This occurred because various Teams features depend on different Microsoft services.

The disruption showed how a single technical failure can impact connected functions without necessarily bringing down an entire application.

Microsoft’s probe centered on issues impacting its cloud infrastructure and the systems in charge of managing service connections.

Reports suggested the incident involved Exchange Online, Microsoft’s cloud-based email and calendar platform. Issues with this infrastructure impacted Outlook and other connected services.

Based on reporting about Microsoft’s investigation, a process in charge of setting up secure connections on affected servers became overloaded. Server availability also dropped after machines failed their automated health checks.

Did a Cyberattack Cause the Microsoft Outage?

Microsoft stated that there was no sign that malicious activity or a cybersecurity attack was behind the outage.

The company concentrated its recovery work on bringing affected servers back online, managing traffic, and implementing infrastructure fixes. As a result, users should not simply assume the incident was a security breach.

Still, in the reports available during the disruption, Microsoft had not offered a full technical explanation of every element of the incident.

Microsoft sought to bring services back online by implementing technical fixes and rerouting traffic away from the affected infrastructure.

The steps it took to recover included:

• Rebooting servers that had been affected.
• Redirecting more traffic away from the problematic infrastructure.
• Halting some deployments.
• Implementing fixes to authentication and routing configurations.
• Watching server stability as services came back online.

At first, several recovery steps yielded only partial improvements. Microsoft later reported favorable results after applying a patch to a portion of its infrastructure. A full recovery, however, was not confirmed right away.

The company also issued an apology for the disruption, recognizing the effect on customers and businesses.

The disruption extended beyond individual users who were having trouble with their email accounts.

Companies, government bodies and financial institutions depend on Microsoft 365 for communication, documents and everyday operations. Once these services go down, even ordinary tasks can be delayed.

Reports indicate that the outage impacted organisations in 74 countries and territories. Certain businesses had difficulty accessing email, while others ran into issues with internal communication and scheduling.

Effect on Indian Airlines

The outage also disrupted the email support services that Indian airlines rely on.

IndiGo, Air India, and Akasa Air released customer advisories regarding disruptions linked to email. They urged passengers to rely on other communication channels for urgent inquiries.

Air India stated that flight operations were unaffected, whereas IndiGo and Akasa Air pointed customers to their available support channels.

This demonstrated how a cloud service issue can impact customer support even while an organisation’s core operations keep running.

The failure of ChatGPT and Claude connectors points to a key problem in the expanding AI ecosystem: AI assistants frequently rely on external platforms to carry out useful workplace tasks.

Today’s AI tools go beyond simply generating text. They link up with email services, calendars, document storage platforms, and workplace applications.

These integrations assist businesses in automating tasks and boosting productivity. Yet they also create dependencies.

For instance, an AI assistant might be available, but it cannot summarise an email held in a service that is down. Likewise, it may be unable to check a calendar when the connected platform cannot respond.

Why Companies Require Backup Systems

The Microsoft outage provides several lessons for organisations that rely on cloud-based applications.

• Keep alternative communication channels available: Businesses should have backup options for urgent communication.
• Arrange alternative access to data: Important documents should stay accessible through approved backup arrangements.
• Examine third-party dependencies: Organisations should know which services depend on the same cloud provider.
• Develop business continuity plans: Teams need clear procedures for operating during technology disruptions.
• Track service status: IT teams should follow official updates from cloud and software providers.

These measures cannot prevent every outage. However, they can lessen operational disruption and help businesses recover more quickly.

Microsoft’s restoration work matters not only for its own customers but also for companies that rely on third-party apps linked to Microsoft 365.

OpenAI and Anthropic also need to keep an eye on their integrations while Microsoft brings its services back online. Whether these connectors remain available hinges on how stable the underlying Microsoft infrastructure is.

The incident also prompts questions about how AI companies can make their integrations more resilient. More and more, businesses expect AI assistants to operate across multiple applications without disruption.

Moving forward, companies may require more robust fallback systems, clearer communication during outages, and improved contingency planning for connected services.

All in all, Microsoft’s large-scale outage shows how a disruption at a single major technology provider can ripple across an entire ecosystem of applications. Although the core AI capabilities of ChatGPT and Claude are separate from their Microsoft 365 connectors, the incident reveals how much modern AI workflows can depend on external cloud infrastructure.