Internet Status — July 25, 2026: Core Infrastructure Healthy, Twilio Carrier Issues Ongoing
No widespread internet disruption is in progress. Core infrastructure — the major cloud regions, CDNs, and public DNS — is operating normally. A few narrower issues remain open, and GitHub is recovering from an unusually busy 24 hours.
**What is active right now**
Twilio is the main open item. Its status page lists SMS delivery delays and failures from US carrier networks to a subset of Twilio short codes (under investigation since July 24), voice call failures to Reliance Jio and other Indian networks (cause identified, fix in progress since July 17), and delayed delivery receipts for Claro Guatemala subscribers. These are carrier-level problems affecting messaging and voice traffic rather than general web access. See [status.twilio.com](status.twilio.com).
Cloudflare's dashboard shows its usual scattering of partial outages at small edge locations, where traffic is automatically re-routed to nearby data centers. This is normal background state for a network of that size and is not a service-wide problem.
**Resolved in the past 72 hours**
GitHub had a difficult stretch and has now fully recovered. Today it reported degraded availability across several GPT models in Copilot between 09:42 and 10:11 UTC, which it attributed to an issue with an upstream model provider, plus a short Actions queueing delay earlier in the morning. On July 24 it saw a critical failure creating pull requests (19:37–20:23 UTC) and a broader disruption touching the API, Issues, Actions, Pages, and Copilot (16:17–17:36 UTC). Separately, an abuse-mitigation change incorrectly blocked some legitimate traffic through GitHub's Central Europe and South America edge locations from 18:45 UTC on July 23 until 11:19 UTC on July 24. All are marked resolved at [githubstatus.com](githubstatus.com).
AWS resolved a connectivity event on the morning of July 24 (roughly 04:40–06:01 PDT). Akamai mitigated an edge delivery issue causing latency and 5xx errors on July 24 at 12:00 UTC, traced to a third-party provider. DigitalOcean cleared a routing problem affecting traffic from India to its NYC region, limited to Airtel subscribers. Sony's PlayStation Network had a multi-hour global outage on July 24 affecting sign-in, multiplayer, and the store; services have been restored.
**A note on unverified reports**
At least one outlet published a story today describing a global, ongoing ChatGPT outage and stating that OpenAI had acknowledged elevated error rates on its dashboard. We could not substantiate this. OpenAI's status page reports all systems operational with no ongoing incidents or affected components, and both chatgpt.com and the OpenAI API responded normally to our checks. GitHub's brief Copilot model degradation this morning does point to a real upstream model-provider hiccup in an overlapping window, but we found no evidence of a broad or continuing ChatGPT outage. Treat that report with caution.
**Everything else checked**
Operating normally: [AWS](health.aws.amazon.com), [Google Cloud](status.cloud.google.com) (no broad severe incidents), [Azure](azure.status.microsoft) (no active events), [Cloudflare](cloudflarestatus.com), [Akamai](akamaistatus.com), [OpenAI](status.openai.com), [Discord](discordstatus.com), [Atlassian](status.atlassian.com), [Slack](slack-status.com), [Zoom](zoomstatus.com), [Vercel](vercel-status.com), [Heroku](status.heroku.com), [DigitalOcean](status.digitalocean.com), [Stripe](stripestatus.com), and Meta. Public DNS resolvers (1.1.1.1, 8.8.8.8, 9.9.9.9) answered in 31–39 ms, and no significant BGP hijacks or route leaks surfaced in our checks.
Two sources could not be read this run: Fastly's status page and API returned HTTP 403, and Downdetector blocked automated access. We saw no independent signal of problems at either provider, but our Fastly coverage today is incomplete.
**Bottom line:** the internet is up. Expect messaging delays if you send SMS to US short codes or place voice calls into India via Twilio; otherwise things are quiet.