Microsoft apologised to enterprise customers and partners on Friday after an 18-hour infrastructure failure crippled its flagship communication and productivity tools across at least 74 countries.
The outage struck email, corporate chats, and artificial intelligence interfaces without warning. Businesses ground to a halt. School administrations scrambled. Government departments lost internal communication lines, while international airports wrestled with check-in snarls. In India, major commercial carriers issued travel advisories warning passengers of check-in and booking delays tied directly to the cloud failure.
The disruption was massive, sudden, and global.
Microsoft ruled out sabotage, hostile nation-state intrusion, and rogue artificial intelligence agents. Instead, the Redmond-based tech giant pointed directly to internal engineering missteps. The company confirmed that a technical defect introduced during an earlier network capacity upgrade flared up immediately following a routine network modification.
We are deeply sorry for the disruption this incident has caused our customers and partners. We know the trust they place in our services and take that responsibility extremely seriously. We are seeing broad recovery and stability across our services. This was a technical issue not related to cyber or AI/agents. The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change.
Microsoft
Engineers bypassed and repaired the underlying code before restoring traffic in phased waves. Microsoft confirmed it validated system stability at every step to prevent traffic surges from crashing recovered nodes. That staggered strategy lengthened the downtime. But it prevented a second cascade.
| Disruption Metric | Peak Impact Recorded | Recovery Benchmark |
|---|---|---|
| Outlook Disruption Reports | 138,058 reports (1:00 PM ET) | Slumped to 34,490 by Thursday evening |
| Microsoft 365 Incident Reports | Tens of thousands logged | Dropped to ~13,000 by 5:47 PM ET |
| Microsoft Teams Reports | Multi-region failure | Tumbled to 763 reports by late evening |
| Affected Jurisdictions | At least 74 nations | Gradual global restoration underway |
The failure reached far beyond standard desktop software. The outage battered Outlook and Microsoft Teams, while also severing external AI connectors utilized by top artificial intelligence developers, including OpenAI and Anthropic.
Real-time monitoring platform Downdetector tracked the acute wave of service failures as corporate workdays opened across North America and Europe. Outlook outage complaints surged past 138,000 at 1:00 PM Eastern Time. By Thursday evening, unresolved complaint volumes dropped back to 34,490 as regional servers reconnected. Reports on Microsoft 365 tapered to roughly 13,000 by 5:47 PM Eastern Time, while Teams issues shrank to 763.
Microsoft engineers are monitoring residual latency across affected regions as internal review teams compile telemetry data. The company announced it will publish a full post-incident report detailing root causes and safeguards once its technical audit concludes.