Emerging ThreatsData Breaches

GitHub Outage Disrupts Global Access to Key Services

Analyst 207||Source: BleepingComputer
Employees sit at desks with concerned expressions, screens displaying errors, in a large office space with a system status…

"is investigating reports of performance problems affecting some of its services," GitHub wrote at 9:40 AM EDT on August 17, 2026 — a line that marked the start of a fast-moving, widespread outage affecting large parts of the platform developers depend on.

Timeline: confirmation, spread, and Copilot impact

GitHub confirmed the outage at 9:40 AM EDT on August 17, 2026 and said it was investigating performance problems affecting some services. Within an hour the incident had expanded: at 10:31 AM EDT GitHub confirmed that Copilot was experiencing degraded availability, extending the outage to the company's AI coding services. GitHub has not disclosed a cause and says its investigation is ongoing.

Which services are degraded and the measured error rates

According to GitHub's status page, the company is seeing error rates of around 20% across its web experience and API traffic. The outage has produced server errors for some users and intermittent failures for others when attempting to load commits, repositories, and Pull Request pages. Several core features developers rely on — API Requests, Actions, Webhooks, Issues, and Pull Requests — are explicitly reported as affected by the incident.

Repository downloads and authentication services hit hardest

The problems are uneven across services. Archive downloads and raw repository content downloads are experiencing error rates of approximately 50%, according to the status reporting — a markedly higher failure rate than the general web and API traffic figures. Authentication-related services are also disrupted: SAML and OIDC authentication, SCIM, and Team Sync are listed as affected, which can prevent organizations from validating or synchronizing user access in affected environments.

CI/CD pipelines, automated workflows, and developer tooling

GitHub Actions is listed as experiencing degraded performance, which the company warned can affect automated builds, tests, deployments, and other workflows that depend on its CI/CD platform. That degradation, combined with issues in API, Webhooks, and Pull Requests, creates a compound impact on end-to-end developer workflows. Copilot — GitHub's AI coding service — was later confirmed to have degraded availability, broadening the incident from platform and CI components into developer-assist tooling.

How developers, enterprises, and CI/CD teams are affected

  • Developers: Many users are encountering server errors when accessing repositories or pull requests and are seeing failures when loading commits and repo pages; archive and raw content downloads are failing for roughly half of attempts.
  • CI/CD teams and build engineers: Degraded performance in GitHub Actions and webhooks can interrupt automated builds, tests and deployment pipelines that rely on GitHub for triggers, artifacts, or configuration.
  • Enterprises and identity managers: With SAML, OIDC, SCIM, and Team Sync affected, organizations that depend on those services for authentication and team provisioning may face sign-on and synchronization failures until the issue is resolved.

Several components remain listed as operational — Git Operations, Packages, Pages, and Codespaces — even as multiple other services report degradation or elevated error rates. GitHub has not provided a root cause and continues to investigate, leaving developers and administrators to monitor the status page and mitigate where possible.

The event underscores the fragility that can follow when central developer platforms experience system-wide errors: a single outage can cascade from simple page loads to authentication flows and automated delivery pipelines. GitHub's public status metrics — 20% error rates for web/API traffic and ~50% for archive/raw downloads — give a numerical snapshot, but the company’s ongoing investigation will be the primary source for confirmation of cause and recovery timing.

Original reporting: BleepingComputer — "Microsoft confirms GitHub is down worldwide"

Artificial IntelligenceCloud ServicesCopilotDeveloper ToolsEmerging ThreatsService DisruptionGithub Outage
Related Intelligence

Continue Reading

Smartphone on a plain surface with blurred cityscape in background.

UNISOC Modem Flaw Enables Remote Code Execution

A newly discovered flaw in UNISOC modem firmware can be exploited to execute arbitrary code with kernel privileges, allowing hackers to gain deep access to Android devices. Simply making a video call to a vulnerable phone can trigger the attack.

Analyst 207
Server room interior with technicians in background and highlighted server components.

China APT Exploits VMware Flaw in Targeted Attacks

A recent investigation revealed that a suspected China-nexus APT group is actively exploiting a critical VMware vCenter vulnerability, CVE-2026-59310, to execute arbitrary code and deploy a backdoor, with ransomware seemingly used as a smokescreen to distract from the underlying intrusion. The attackers' true intentions appear to go beyond mere ransomware deployment.

Analyst 207
Smartphone on cluttered office desk with cityscape background through window.

Unisoc Exploit Chain Grants Attackers Full Android Kernel Access

Security researchers have uncovered a two-stage exploit chain that can give attackers full access to the Android kernel on devices using Unisoc modem firmware, and alarmingly, the vendor has remained unresponsive to disclosure efforts. This chain can be triggered by a simple malformed video call, putting countless devices at risk.

Analyst 207
Rows of computer servers and storage equipment in a brightly-lit data center with concerned businesspeople in the background.

Clop Ransomware Targets GE, Philips in Data Theft Attacks

Major companies like Philips, General Electric, and Shell are investigating claims by the Clop ransomware gang that their systems were breached, with Philips confirming a contained breach of an internal server that didn't affect customers. The incidents are a stark reminder of the growing threat of ransomware attacks on businesses.

Analyst 207