GitHub outage history

An incident briefly affected GitHub Actions workflows and related services, causing degraded performance for a period beginning at Aug 18, 2026 9:41 AM UTC when GitHub reported investigating reports of impacted performance for some services. The issue was resolved by Aug 18, 2026 10:26 AM UTC as affected services returned to normal. GitHub's status updates described the investigation and eventual restoration, marking the end of the disruption after about 45 minutes for users relying on Actions-enabled workflows.

GitHub.com experienced a broad outage beginning around 1:41 PM UTC, with about 20% error rates across core experiences such as Pull Requests, Issues, API traffic, and other web surfaces. Users reported failures across workflows including code reviews and Copilot access in some editors. StatusGator detected the issue at Aug 17, 2026 12:47 PM UTC, about 53 minutes before GitHub posted its first incident update at 1:41 PM UTC. The disruption appeared global, affecting users worldwide rather than being confined to a single region.

GitHub reported a brief disruption to GHEC Team Sync that degraded provisioning for enterprises with personal accounts, causing delays when syncing teams to IdP groups. The issue began at 4:26 PM UTC and was resolved by 6:31 PM UTC.

GitHub identified the cause and began implementing a mitigation, with a progress update planned for 8:00 PM UTC.

Users relying on Copilot experienced degraded availability as the Fable 5 model in Copilot faced issues from an upstream provider. The impact extended to related GitHub features, with a degradation noted across Git operations, pull requests, issues, and the package registry, and an associated disruption to webhooks between 2:32 PM UTC and 2:46 PM UTC. The situation was acknowledged in updates starting around 2:52 PM UTC, as engineers worked with the upstream provider to isolate and mitigate the root cause.

Users experienced degraded Webhooks and related GitHub services, with pushes, issues, pull requests, and the package registry briefly affected. The degradation began around 2:32 PM UTC, and the issue was acknowledged as a Webhooks degradation by 2:46 PM UTC. Separately, Copilot users faced degraded availability for the Fable 5 model across Copilot products and IDE surfaces due to an upstream model provider issue.

StatusGator detected issues at Aug 12, 2026 8:29 PM UTC—an Early Warning Signal—indicating the GitHub website was inaccessible due to server connection errors, which persisted until 10:00 PM UTC. The outage primarily affected user authentication and the ability to download Release assets, with many users reporting login failures and failed asset retrieval during the window. Website accessibility was also impacted in the early phase.

Users on GitHub.com experienced degraded performance for Pull Requests and Issues, with some pages failing to load and returning 500 errors. The disruption affected how developers view and interact with pull requests and issue discussions, slowing collaboration. The incident began at 4:16 PM UTC and was resolved by 4:36 PM UTC, lasting about twenty minutes.

GitHub's status updates indicated an investigation into the root cause, with the update noting that the issue appeared related to a database infrastructure issue. The teams worked to restore normal operation and by 4:36 PM UTC services were returned to normal for PRs and Issues.

The GraphQL API requests on GitHub began to degrade at Aug 11, 2026 2:51 PM UTC, with users experiencing slower responses and timeouts. Some Copilot users reported failures completing simple requests, illustrating the impact on GraphQL consumers. The incident was initially characterized as degraded performance affecting the API Requests component.

A second update at Aug 11, 2026 3:31 PM UTC noted a small increase in error rates and ongoing capacity work, with GitHub investigating and working to increase capacity. By Aug 11, 2026 4:51 PM UTC the API Requests component had returned to normal, and services were back to typical performance, marking the end of the incident. The overall impact appeared limited to GraphQL API users and those relying on Copilot, with no broader outage.

GitHub reported a brief disruption affecting access to certain models via GitHub Copilot, causing degraded performance for some GitHub services. The incident began at 8:31 PM UTC and was resolved by 8:41 PM UTC as GitHub investigated reports of impaired performance and confirmed a Copilot-related issue. Services returned to normal within about ten minutes, with no lasting impact.

GitHub experienced a brief disruption affecting the creation of fine-grained Personal Access Tokens. During the incident, attempts to create a new token redirected users to the token overview page without an error message and no token was created. The disruption began at 6:07 PM UTC and was resolved by 6:26 PM UTC as GitHub's status updates indicated the underlying issue was addressed and services returned to normal.