Table of Contents
Why are ChatGPT and Microsoft Defender not working right now?
Tuesday, December 2, 2025, proved difficult for IT administrators and digital professionals globally. Multiple critical services experienced significant instability simultaneously. If your organization struggled with communication or security management yesterday, you were not alone. Three distinct service disruptions occurred involving Microsoft Exchange Online, the Microsoft Defender Portal, and OpenAI’s ChatGPT.
Here is a technical breakdown of what happened, the specific error codes involved, and the current status of these services.
Exchange Online: Connectivity Failures (Ongoing Investigation)
Microsoft is actively investigating a connectivity issue affecting Exchange Online. This disruption prevents users from accessing mailboxes through specific connection protocols. While the service is not completely offline, the instability is creating intermittent access failures.
Technical Details:
- Advisory ID: EX1191546
- Impact: Users face errors when attempting to sync or access email via Outlook, mobile clients, or Outlook on the Web (OWA).
- Current Status: Microsoft engineers are analyzing the root cause to determine the full scope of impact on organizational tenants.
- Advisory Note: Unlike a total blackout, this issue manifests as specific connection method failures. Your desktop app might fail while the web app works, or vice versa. IT teams should monitor the Microsoft 365 Service Health Dashboard for real-time updates on this specific ID.
Microsoft Defender Portal: Traffic-Induced Glitches (Resolved)
Security teams faced a frightening scenario yesterday morning when the Microsoft Defender Portal became inaccessible. For cybersecurity professionals, losing access to the management pane prevents the investigation of alerts and the configuration of threat policies.
Timeline of Events:
- Issue ID: DZ1191468
- The Cause: Microsoft identified an unexpected spike in network traffic directed at the Defender infrastructure. This surge overwhelmed the system, causing timeouts for administrators trying to log in.
- The Fix: Engineers applied traffic mitigation strategies to handle the load.
- Resolution: Telemetry data confirms the service has stabilized. Access to the Defender admin center is fully restored. If you still see errors, clear your browser cache and attempt to reconnect.
ChatGPT: Global Service Interruption
Coinciding with the Microsoft issues, OpenAI’s ChatGPT suffered a widespread outage. This disruption halted workflows for millions of users relying on AI for coding, writing, and data analysis.
The Impact:
- Users attempting to load the interface received generic internal server error messages.
- History and previous chats became temporarily unavailable.
- Unlike the Microsoft issues, this outage appeared to be a total service denial rather than a partial degradation. Service has since returned to normal operational levels.
Advisor Recommendations
When multiple cloud dependencies fail simultaneously, it highlights the fragility of modern digital ecosystems. For the Exchange Online issue (EX1191546), which may still be affecting tenants, instruct your users to attempt alternative access methods. If the desktop client fails, guide them to the web browser version.
Document these outages internally. Tracking vendor reliability (SLA monitoring) is essential for year-end vendor assessments and compliance reporting.