GitHub down for most of morning
Archived — this story has rotated out of today’s deck. It is kept here in full.
The gist
GitHub went down for most of Monday morning, hitting APIs, pull requests, and Copilot.
The outage disrupted developers worldwide, with error rates up to 50% on downloads.
Background
GitHub, a Microsoft-owned platform with 225 million users, experienced a widespread outage on Monday, August 17, 2026. The issue started around 6:40 a.m. Pacific and affected many core services, including the website, API, pull requests, webhooks, and Copilot. GitHub's status page reported high error rates, with web traffic at about 20% and downloads at about 50%. The company implemented a fix and showed signs of recovery by midday, but the root cause was not disclosed.
How it unfolded
- Aug 17, 2026 06:40 PTGitHub outage begins, affecting multiple services.
- Aug 17, 2026 09:30 PTGitHub says it found the source of the problem.
- Aug 17, 2026 10:00 PTGitHub brings the outage under control, with services recovering.
- Aug 17, 2026 12:45 ETOutage reports drop to a few hundred, GitHub reports 'strong signs of recovery'.
Who’s saying what
- Official
- GitHub acknowledged high error rates and said it was performing mitigations and monitoring for improvement.
- Analysts
- The outage is a setback for Microsoft's coding platform, disrupting developers worldwide.
Still unverified
The root cause of the outage has not been officially disclosed; some reports suggest it may be related to infrastructure issues, but this is unconfirmed.