The disruption affected Microsoft 365 integrations used by ChatGPT and Claude, causing errors across email, calendar and workplace tools as Microsoft worked to restore services.
A widespread Microsoft 365 outage has disrupted artificial intelligence tools operated by OpenAI and Anthropic, affecting integrations used to access workplace applications including Outlook, Teams, SharePoint and calendar services.
The disruption, reported on November 13, 2026, affected Microsoft's enterprise applications and extended to third-party AI platforms that rely on its services. Both OpenAI and Anthropic acknowledged problems with their Microsoft 365 connectors through their respective service-status pages.
OpenAI reported elevated errors affecting Microsoft 365 integrations within ChatGPT. According to the company's status update, users attempting to establish new connections to Microsoft services could experience failures, while existing connected tools might not function as expected.
The company began investigating the issue on November 12, identifying degraded performance within ChatGPT's connectors and applications. At the time of the reported update, OpenAI had not disclosed a specific cause or confirmed a resolution.
Anthropic also reported disruptions affecting Microsoft 365 integrations used by its Claude AI assistant.
The company said users experienced slow responses and elevated errors when accessing email, SharePoint, Teams chat and calendar functions through its connector. Some users were also unable to complete new connections.
Anthropic subsequently identified an issue involving an upstream service provider, reporting that most requests involving mail, calendar and search tools were failing.
The disruption also affected Claude for Outlook, where users encountered problems performing email and calendar actions. Some users were unable to open the Outlook add-in.
Anthropic later attributed the problems to a Microsoft 365 service incident and said it was coordinating with the provider to restore functionality.
While Claude's Microsoft-dependent features experienced degraded performance, Anthropic listed its Claude API, Claude Code and Claude Console services as operational.
Microsoft, meanwhile, reported progress in addressing the broader outage, which had affected Outlook, Teams, Copilot and other Microsoft 365 applications.
The company said it had applied a configuration patch targeting authentication and routing infrastructure, with initial monitoring showing improvements in server stability.
However, Microsoft had not confirmed a timeline for complete service restoration at the time of the update.
The incident highlights the operational dependencies between enterprise AI assistants and the cloud-based productivity platforms they connect with.
As organisations increasingly use AI tools to retrieve emails, manage calendars, search documents and support workplace activities, disruptions to underlying services can affect multiple applications simultaneously.
The outage demonstrates how problems within shared enterprise infrastructure can extend beyond the original service provider, affecting AI-powered workflows used by businesses.
At the time of the reported updates, Microsoft and the affected AI companies were continuing restoration efforts, with full recovery yet to be confirmed.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.