Internet Status — July 19, 2026: GitHub Actions Outage Ongoing; Brief Facebook/Instagram Disruption Recovered

The internet's core infrastructure is healthy today — no widespread outage is in progress — but a few provider-level incidents are active or recently recovered.

GitHub is experiencing an ongoing incident affecting GitHub Actions and API requests (githubstatus.com). Actions is in a partial outage with critical impact and API requests show degraded performance; GitHub is still investigating as of 02:23 UTC. The disruption has knock-on effects: OpenAI reports elevated errors in ChatGPT and Codex workflows that depend on GitHub (status.openai.com), and CI/CD pipelines relying on Actions may see failures or delays.

Earlier today, Facebook and Instagram users worldwide reported a brief outage between roughly 3:44 and 5:02 a.m. ET, with more than 23,000 Facebook and 18,000 Instagram problem reports in the US alone before report volumes dropped sharply (cnn.com/2026/07/19/world/facebook-instagram-outages-intl). NetBlocks confirmed the disruption was platform-level, not a country-level internet issue. Service has recovered.

Elsewhere, Akamai has two incidents in the monitoring/recovery phase — edge delivery issues in India and an Application Load Balancer reporting issue — both trending toward normal (akamaistatus.com). Cloudflare shows only a minor degradation affecting two Workers AI models, and Twilio lists routine regional carrier-level SMS/voice delays.

Recently resolved (past 72 hours): AWS CloudFront suffered a roughly 3.5-hour outage on July 16 tied to a configuration loading failure in its VPC Origins feature, affecting sites including Hugging Face, Canvas, and Ubiquiti — fully recovered. GitHub also resolved a July 17 incident in which about 39% of REST API requests failed.

All clear at AWS, Google Cloud, Azure, Azure DevOps, Discord, Atlassian, Slack, Zoom, Vercel, Heroku, DigitalOcean, and Stripe. Crowd-sourced outage trackers show no other elevated reports.