Microsoft Azure OpenAI

Impact Statement: Starting at 23:54 UTC on 26 September 2025, customers in Switzerland North may experience service unavailability or degraded performances for resources hosted in the region. Virtual Machines may have shutdown to preserve data integrity. Current Status: We were alerted to this issue by our telemetry informing us in a significant drop in traffic. It was discovered that a recent deployment introduced a malformed prefix in one of the certificates used for connection authorization. We have pinpointed the deployment error involving the certificate prefix and are rolling back the faulty deployment to restore normal traffic flow and service availability.We are seeing success and will continue to monitor traffic and service stability to ensure full recovery, at this stage we expect to fully mitigate by 05:30 UTC.

Impact Statement: Starting at 09:00 UTC on 27 January 2026, customers using Azure OpenAI Service in Sweden Central may experience intermittent availability issues. This issue may have also affected Cognitive Services dependent on OpenAI in this region.Workarounds: As temporary workarounds, if your architecture allows, we recommend failing over to another Azure region (such as West Europe, UK South, or East US) and avoiding the deployment of new Azure OpenAI resources in Sweden Central until the issue is fully resolved.Current Status: Our engineering teams are investigating this issue for Azure OpenAI in Sweden Central and it experienced an outage due to a platform-level failure in the AKS cluster hosting internal resource manager (IRM). We have identified that contributing factors include memory pressure causing widespread out-of-memory terminations, IRM crash-looping, network instability, Redis pool exhaustion, and failed Cosmos database (DB) fallbacks—resulting in 500/503 errors for customers. The outage has affected major models (GPT-5.0, GPT-5.2). Mitigation efforts like scaling nodes, network troubleshooting, and rollbacks have shown some improvement, and we are monitoring progress to ensure sustained stability. Investigations and troubleshooting are ongoing. We will provide another update within 1 hour or sooner as new information becomes available.