Microsoft 365 outage leaves small subset affected into third day; ChatGPT Work recovers after five-hour disruption
Two unrelated workplace-service disruptions affected Microsoft 365 and OpenAI users beginning on Monday, August 31. Microsoft said on Wednesday that Microsoft 365 availability remained above 99%, although a small subset of users could still experience impact as recovery work continued into a third calendar day. Separately, OpenAI said ChatGPT Work fully recovered after elevated errors and latency disrupted tasks for about five hours and 24 minutes.
Microsoft 365 outage reaches third day with limited residual impact
Microsoft first tracked the Exchange Online disruption under incident EX1464935 before opening the broader Microsoft 365 incident MO1465074. Reported symptoms included delayed or failed email delivery, mailbox search failures or incomplete results, authentication errors, administrative problems and intermittent mailbox operations. TechCrunch reported on Tuesday that Microsoft was still expanding mitigation while service availability improved.
Microsoft later said the same underlying problem affected functions in OneDrive for Business, SharePoint Online, Teams, Microsoft Purview, Microsoft Defender XDR and Microsoft 365 Copilot. The company attributed the disruption to an issue in a core authentication configuration used by multiple Microsoft 365 services. Its mitigation reapplied authentication components and restarted targeted infrastructure.
Recovery was staged rather than immediate. Computerworld reported that much Exchange Online mail flow recovered while search and related functions remained affected into a second day. BleepingComputer later reported that Microsoft declared mail flow and search restored on September 1, then said on September 2 that availability was stable above 99% and most users should no longer see impact. Microsoft nevertheless continued work for a small subset that could still encounter problems.
Those updates also show why a single recovery percentage cannot describe every user’s experience. Microsoft’s listed scenarios covered different functions, from message delivery and mailbox search to authentication and administration. The company continued balancing its fixes across the remaining infrastructure after broad availability improved, so an organization could see normal service in one function while a limited issue persisted elsewhere.
The evidence does not support calling this a total Microsoft outage. Microsoft’s public status page identified degradation in Microsoft 365 business and enterprise services, while consumer Microsoft products and Azure were shown as operational.
ChatGPT Work recovers after elevated errors and latency
OpenAI opened a separate ChatGPT Work incident at 15:04 UTC on August 31. The company said users across multiple subscription plans could be unable to start or continue tasks, with Plus users particularly affected early in the incident. OpenAI classified the impact as major and identified elevated errors and latency as the confirmed symptoms.
CNBC reported the interruption alongside the unrelated Microsoft problems, while the USA Today Network separately covered user difficulties and OpenAI’s response. OpenAI applied a mitigation at 20:01 UTC and declared all affected services fully recovered at 20:28 UTC.
OpenAI’s incident record progressed from investigation to mitigation and recovery within the same day. Its final all-clear makes this a completed incident, unlike Microsoft’s continuing work for a small residual group. The five-hour duration therefore applies to the recorded ChatGPT Work event and should not be extended to other OpenAI services.
The confirmed scope was ChatGPT Work rather than every ChatGPT, API or Codex function. There is also no evidence that OpenAI’s incident was caused by or connected to Microsoft’s authentication-configuration problem.
Track service recovery without overstating the evidence
Outage scope and restoration status can change quickly. Readers can follow the Microsoft Outlook service page, check the ChatGPT service page, or review the Status Is Down real-time dashboard for current signals. Users can also create an account to monitor selected services and receive alerts when fresh disruptions emerge.
Sources
- Microsoft Service Health: Microsoft service health
- TechCrunch: Microsoft 365 outage drags on, but things are improving
- BleepingComputer: Massive Microsoft 365 outage causes auth issues, service failures
- Computerworld: Microsoft 365 outage enters second day as search disruptions persist
- OpenAI Status: ChatGPT Work experiencing elevated errors
- CNBC: Microsoft Outlook and OpenAI’s ChatGPT Work are experiencing user outages
- USA Today Network: Is ChatGPT down? What users should know about reported outage
Image credit: Fizkes | Dreamstime.com — https://www.dreamstime.com/african-man-using-laptop-application-failure-message-scr-computer-error-failure-concept-african-man-using-laptop-image109505513