The GitHub platform, which belongs to Microsoft, has encountered a major outage that has affected virtually all key services. The problems have impacted the web interface, the programmatic interface, pull requests, webhooks, and the Copilot AI assistant — in other words, the tools that support the daily work of millions of developers around the world. The platform officially recorded the incident at 13:40 UTC on Monday, announcing the start of an investigation into complaints about a decline in the performance of several services at once. Within a few minutes, API requests and Actions received degradation status; subsequently, webhooks, tasks, and pull requests were added to the list.

A complete shutdown of the platform did not occur, but the scale of the problems turned out to be quite significant. According to GitHub itself, the error rate in the web interface and API traffic reached approximately twenty percent, and when downloading archives and making requests to the raw content of repositories, this figure approached fifty percent. Later, the GitHub Pages service also began to experience performance degradation, while Copilot was added to the list of affected components even earlier. At the same time, basic Git operations, packages, and Codespaces continued to operate normally at the time of publication. The breadth of the failure clearly demonstrates how far GitHub has moved away from simply hosting source code. Failures in related components can halt the entire development pipeline in a company, so such incidents hit the business harder than it might seem based on formal availability figures.

Questions about the platform’s reliability have been accumulating for more than a month. In July alone, GitHub recorded eight incidents involving service degradation, whereas there were nine in May and six in June. A little over a week ago, GitHub Actions experienced a particularly prolonged outage, during which workflow runs failed with errors or hung in the queue for hours. Later, the platform itself called that incident unacceptable and promised to accelerate work on isolation and fault tolerance. At the same time, a large‑scale migration of infrastructure to the Azure cloud is ongoing; its goal is to reduce the number of common failure points. Against this backdrop, Monday’s outage looks particularly unfortunate, as it affected not just one isolated component, but multiple critical systems at once. GitHub has not yet disclosed the cause of the latest outage, although jokes have already started circulating in the professional community that the culprit was the volume of junk code generated by neural networks.