GitHub reported that its core services are back online after a major outage on Monday, yet AI coding assistant Copilot remains unstable. The disruption began around 9:40 a.m. ET, spreading to the website, code review tools, automated testing, deployment systems, and Copilot. By 11:10 a.m., most GitHub services were either down or performing poorly, with error rates near 20 % for web and API traffic and up to 50 % for repository content and archive downloads.

The company identified the problematic component, applied a fix, and announced at 2:23 p.m. ET that Git Operations degradation was mitigated, though it continues to monitor stability. Copilot is still under investigation as engineers work to restore AI‑assisted coding reliability. Recent capacity planning documents revealed GitHub had originally targeted a 10x increase in infrastructure, then realized it needed a 30x scale to handle growing workloads. The incident follows several recent service problems, including issues with Actions, API requests, login, and Pull Requests, highlighting increasing reliability challenges.

The exact root cause remains undisclosed, and the outage status page stays open.