Breaking news. What started as loose complaints on social networks was confirmed minutes later: GitHub, the platform that holds the code of half the internet, is down — or at least seriously injured — worldwide.
The official status page itself, githubstatus.com, admits it in red as of 13:40 UTC today: “Partial System Outage”, with the incident still in investigating status — that is, not even GitHub knows yet when this ends.
What’s broken right now
It is not an isolated service. According to the incident report itself, there is a critical and simultaneous drop in:
- API Requests — out of order
- Issues — out of order
- Actions — out of order (goodbye CI/CD pipelines)
- Pull Requests — out of order
- Copilot — out of service
And it doesn’t end there: Pages, Git Operations and Webhooks are working “halfway”, in a degraded state. GitHub itself recognizes error rates of ~20% on web and API traffic, and up to ~50% errors when downloading files or entire repositories.
Translated: if today your git push hung, if Copilot stopped autocompleting mid-sentence, or if an Actions pipeline didn’t start — it’s not your connection. It’s GitHub.
Why is it so scary?
Because GitHub is not “just another app.” It is the backbone of how software is built around the world: from the team of a startup in Quito to the updates of critical systems of banks and governments pass, at some point in the chain, through GitHub. When GitHub coughs, thousands of development teams — and their releases, their builds, their deploys today — literally come to a standstill.
To gauge the magnitude: GitHub hosts more than 200 million repositories and is used by more than 100 million developers worldwide. Companies like Microsoft, Google, Amazon, Netflix, and virtually every tech startup on the planet depend on GitHub for their daily workflow. When the platform goes down, the impact isn’t just “I can’t push” — it’s that security updates don’t get distributed, deploy pipelines don’t run, code reviews stall, and releases planned for today are postponed indefinitely.
Real impact: what stopped when GitHub went down
The interruption is not abstract. There are concrete impacts that affect teams and companies around the world:
CI/CD pipelines stopped. GitHub Actions is the engine behind thousands of continuous integration and continuous deployment pipelines. When Actions goes down, teams can’t run automated tests, generate production builds, or deploy new versions. For companies that deploy daily or even multiple times a day, this means that new code accumulates without being able to be released.
Pull Requests blocked. Code reviews — the process by which one developer reviews another’s work before merging it — depend on Pull Requests. Without PRs, the development workflow grinds to a halt. Teams that rely on code reviews as a quality gate cannot move forward.
Copilot unusable. For the thousands of developers who have incorporated Copilot into their daily workflow, losing the AI tool is like losing an assistant that significantly reduces time on repetitive tasks. It is not critical to the survival of the project, but it does impact daily productivity in a measurable way.
Issues and documentation inaccessible. Repositories that use GitHub Issues as a bug and feature tracking system lose visibility over their backlog. Support teams responding to public Issues cannot access the information.
Pages out of service. Thousands of static websites hosted on GitHub Pages go down simultaneously — project documentation, portfolios, technical blogs and even startup landing pages.
What does GitHub say?
For now, little and cautiously: they recognize the problem, say they are “performing mitigations” and promise updates as they progress. There is, as yet, no confirmed root cause or estimated time for resolution.
The lack of detailed communication is, in itself, a problem. Development teams need to know not only that there is an incident, but: what is the exact scope? Which services are affected and which are not? Is there data loss or just service degradation? What is the probable cause? Every minute without data is a minute in which teams make decisions without data — and many of those decisions (temporarily switching to GitLab, pausing releases, notifying customers) have a real cost.
What to do in the meantime?
If you work with GitHub today, here are concrete recommendations:
1. Save your work locally. If you were about to push, make a local commit and save it. Don’t lose pending work in case the incident spreads.
2. Avoid relying on Actions for anything urgent. If you have a pipeline that needs to run today for a release, consider running tests manually locally or on another platform temporarily.
3. Be patient with Copilot. If Copilot stopped working, it’s not a license issue — it’s part of the incident. It will return when GitHub restores the service.
4. Communicate with your team. If you work as a team, advise that delays in deploys or reviews are due to the GitHub incident, not due to lack of production.
5. Don’t try to reposition or hot migrate. The temptation to migrate to GitLab or Bitbucket during an outage is understandable, but doing so under pressure and without planning often creates more problems than it solves. Evaluate that calmly after GitHub recovers.
6. Monitors the status page. githubstatus.com is updated with every change in the status of services. Refresh every 15-20 minutes to stay on top of things.
The underlying debate: dependence on a single supplier
This incident reignites a conversation that the development world has been having for years: is it healthy for so much critical software infrastructure to depend on a single platform? GitHub, which has been owned by Microsoft since 2018, is currently the de facto standard for source code hosting. But the concentration creates a systemic risk: when GitHub goes down, the entire global software development ecosystem goes down.
The alternatives exist — GitLab, Bitbucket, Codeberg — but the inertia is enormous. Migrating a repository is relatively simple; migrating a team’s work culture (issues, PRs, Actions, Copilot) is another issue. And without an alternative that offers the same level of integration and ecosystem, many companies simply absorb the cost of outages and wait.
The lesson for Ecuadorian companies is clear: if your business depends on development tools hosted in the cloud, you need a documented plan B. It doesn’t have to be a full migration — it can be simply knowing how to run your critical pipelines manually or having a mirror of the code on another platform for emergencies.
Previous GitHub Incidents
This is not the first significant GitHub outage. The platform has had several notable incidents in recent years:
- 2021: A DDoS attack affected access to GitHub for several hours.
- 2023: A 4-hour outage left millions of developers without access to repositories and PRs.
- 2024: An authentication issue prevented access to private repositories during business hours in Asia and Europe.
Each incident reinforces the same lesson: availability is no guarantee, and teams must design their workflows assuming that tools can fail.
Incident update
This note will be updated when GitHub confirms resolution of the incident and provides a root cause. If you need real-time information, please consult githubstatus.com directly.
Source: githubstatus.com — active incident at the time of this publication.