Minor incident
Mews: AI tooling degraded due to Claude and Azure Foundry outage
Sep 29, 2026, 15:02 UTC
started
08:06 UTC
resolved September 30, 2026
17 h 04 min
duration
Affected: Mews Operations
Current status of Mews: Operational — live Mews status
Timeline
- Sep 29, 2026, 15:02 UTCmonitoringWe are aware of an issue affecting AI tooling. The cause is an outage at our AI provider and an Azure outage in the Sweden region. AI-powered features may be slow, return errors, or be unavailable. Core Mews functionality is not affected.
- Sep 30, 2026, 08:06 UTCresolvedThe impact from the issue is over as of Tuesday 29th 8PM utc
- Oct 7, 2026, 08:10 UTCpostmortem**Problem** On 2026-09-29T14:30:06Z, AI-powered features in Mews — including AI guest messaging, AI property guidelines, and smart task suggestions — became intermittently unavailable or degraded. This was caused by an outage affecting Azure OpenAI and Azure AI Foundry services in the Sweden Central region, combined with a concurrent degradation of Anthropic's Claude service. Core Mews functionality was not affected. **Action** Our team detected the issue promptly and confirmed that AI features were unavailable due to the upstream outage. We published a status page update to inform customers, monitored the situation throughout the afternoon, and confirmed full recovery on 2026-09-29T19:06:00Z once Azure services returned to normal. **Causes** The outage was caused by infrastructure issues at Microsoft Azure affecting their OpenAI and AI Foundry services in the Sweden Central region. Azure reported that a backend service responsible for managing request metadata experienced timeouts when reading from dependent database and caching layers, resulting in intermittent failures and increased latency for customers in that region. Anthropic's Claude service also experienced concurrent degradation. **Solutions** The issue was resolved by Microsoft Azure restoring normal operations in the Sweden Central region. On our side, we are working on adding in-product indicators so customers are clearly informed when AI-powered features are temporarily unavailable, and we are reviewing our integration points to ensure features recover automatically once upstream services are restored.
Updates as published by Mews (original).
Other services with incidents at the same time
This was part of a wider event: 18 services reported incidents within the hour.
- Ideogram 15:59 UTC · Outage on v3 endpoints
- Resideo 15:59 UTC · First Alert Application-Water Leak Detector Status
- Sage 15:57 UTC · Brightpearl - Partial Outage
- ThreatLocker 15:49 UTC · Portal Slowness
- Qualys 15:46 UTC · US Platform 3: Delay in data processing to the UI for Portal modules (IM-12887)
- Upstash 15:36 UTC · Brief connection proxy degradation in us-east-1
- CloudLinux 15:33 UTC · dnf update aborts on CL9 + EPEL: ea-php81-85-php-gd and alt-php GD require libavif.so.15, EPEL 9 libavif 1.1.1 clashes with alt-common libavif/svt-av1-libs
- Thomson Reuters Legal 15:26 UTC · CoCounsel New Experience Degraded Performance
- SU ITS 15:16 UTC · Qualtrics Service Disruption
- Pitt Digital Systems 15:09 UTC · Vendor Issue Affecting Claude for Education Logins
- GoTo 15:03 UTC · GoTo Meeting, Webinar or Training - Webcam or screen share issues
- uchicago 14:56 UTC · Claude Partial Outage
- Cielo 14:53 UTC · Indisponibilidade parcial nos fluxos de solicitação (maquininhas) e produtos (serviços) de logística através do site.
- Nextech 14:53 UTC · Performance Issue
- SU ITS 14:50 UTC · Claude.ai outage
More Mews incidents
- October 22, 2026 An issue with sending inventory updates to Booking.com
- October 21, 2026 Critical 0 day PHP vulnerability
- October 19, 2026 Downgraded Performance - Delayed emails
- October 19, 2026 Delays in accounting ledgers and trial balance updates
- October 18, 2026 Issue with check in flow in Mews
- October 15, 2026 Booking.com reservations delayed