Internet Status — July 27, 2026: Scattered Provider Incidents, No Internet-Wide Outage
No internet-wide disruption is underway. Core cloud, CDN, and routing infrastructure is healthy, but a handful of individual providers are working through degradations.
What's active right now:
• OpenAI — An incident affecting ChatGPT conversations with elevated error rates opened late on July 27. Mitigation has been applied and OpenAI is monitoring recovery. This follows a cluster of shorter error-rate incidents on July 24–25 that were all resolved. (status.openai.com)
• Zoom — Degraded performance across several features, including offline meeting transcription in the US region, Zoom AI Companion components, My Notes, and Zoom Phone devices. Zoom says root cause is identified and a fix is in progress. Core meetings remain usable. (zoomstatus.com)
• Twilio — Regional carrier-level problems rather than platform failures: SMS delays and failures to Mascom and Orange subscribers in Botswana, voice call failures to Claro in Guatemala, and ongoing voice delivery failures to Reliance Jio in India. (status.twilio.com)
• Cloudflare — The Dallas (DFW) data center is in a scheduled maintenance window (06:00–12:00 UTC), with traffic rerouted and a possible slight latency increase regionally. This is planned work, not an incident. (cloudflarestatus.com)
Core infrastructure all-clear: Amazon Web Services, Google Cloud, and Microsoft Azure report no active events. Akamai, Fastly, GitHub, Discord, Slack, Atlassian, Vercel, Heroku, DigitalOcean, and Stripe are all operational. We found no evidence of BGP hijacks or route leaks, no DNS resolver problems, and no new submarine cable faults affecting global connectivity.
Resolved in the last 72 hours:
• Akamai mitigated a Bot Manager false-positive issue at 01:30 UTC on July 27; normal operation has resumed.
• GitHub resolved degraded GraphQL API performance earlier on July 27, plus Actions run failures on July 25 and a broader multi-service degradation on July 24.
• Stripe cleared elevated API errors on July 24.
• DigitalOcean traced a July 24 connectivity issue for some India-based users reaching NYC resources to an internal Airtel network problem, not DigitalOcean infrastructure.
Consumer-level reports: Crowd-sourced trackers show localized ISP complaint spikes today — Spectrum in the United States beginning around 3:19 a.m. Eastern, Vodafone in Australia, and Biznet in Indonesia. These are regional access issues affecting individual providers' subscribers and are not connected to a broader infrastructure failure.
One note on sourcing: the AWS public RSS feed remains stale, with its newest entries dated April 2026. Our AWS assessment relies on the health dashboard and corroborating signals rather than that feed.
Bottom line: The internet is up. If you're seeing problems today, the most likely explanations are your own ISP, or one of the specific services listed above.