Developers around the world lost most of a working day on Monday when GitHub, the platform that hosts more code than any other, suffered a multi-hour global outage. Errors began at around 9:40 AM ET (13:40 UTC), and the disruption rippled through the core of the modern software development workflow: pull requests that would not merge, CI/CD pipelines stalled by Actions, unreliable issue tracking, and Copilot — GitHub's AI coding assistant — degraded alongside them.
According to GitHub's status page, error rates reached roughly 20% for general web experiences and API traffic, while archive downloads and raw repository content downloads failed at an alarming approximate 50% rate — effectively halving the reliability of those operations. Enterprise customers faced additional pain: SAML and OIDC authentication, SCIM provisioning and Team Sync were all flagged as impacted, threatening single sign-on access for organizations on GitHub Enterprise Cloud.
The incident unfolded methodically. GitHub first acknowledged "impacted performance" at 13:40 UTC, then confirmed degraded API requests, Actions and Webhooks within the hour; Issues and Pull Requests were added shortly after, and Copilot joined the list around 14:31 UTC. Independent monitoring service Downdetector showed user complaints spiking near 13:30 UTC, and Microsoft confirmed the platform was experiencing worldwide issues.
There was some relief in the afternoon: at 1:34 PM ET GitHub said it had pinpointed "the problematic component" and taken corrective action, restoring webhooks, API requests and pull requests. But Copilot and the GitHub website remained troubled until 5:15 PM ET, when the company gave the all-clear.
GitHub has not yet said what caused the outage, promising a "detailed root cause analysis" once complete. The incident is at least the second major Microsoft platform outage in recent weeks, after Xbox services were knocked offline for several hours in late July — a pattern that has fueled criticism that the company's infrastructure investment is not keeping pace with its platforms' centrality. The outage landed just months after GitHub rolled out a more transparent status-reporting framework in April 2026, which gave developers unusually detailed visibility into exactly which services were suffering — even if it could not make the fixes come faster.
