No widespread internet outage is under way this morning (checked around 10:30 UTC). AWS, Google Cloud, Azure, Azure DevOps, GitHub, OpenAI, Anthropic, Discord, Slack, Atlassian, Vercel, Heroku, DigitalOcean, Stripe and the AI platforms we track all report normal operation. A few regional incidents are still open.
Ongoing:
Akamai: Edge Delivery issues in Canada. Opened 2026-09-28 at 16:22 UTC and still marked "investigating" at the last update (23:01 UTC). Akamai has not described the scope. Akamai also still lists origin connectivity errors in APAC, which it attributes to subsea cable repairs by an upstream carrier, and higher latency on routes between Singapore, Hong Kong and Japan during planned carrier maintenance that runs until October 3. (akamaistatus.com)Cloudflare: "Network Performance Degradation, Asia-Pacific" has been open since September 23. Cloudflare says intermittent connectivity problems affect a small subset of customers because regional capacity is badly degraded. This matches the subsea cable work Akamai reports. (cloudflarestatus.com)Twilio: minor carrier-specific messaging problems, including SMS delivery receipt delays to CTBC Brazil (opened September 28) and older MMS delays in Australia and to GCI in the US. (status.twilio.com)Zoom: a partial degradation, open since September 25, that takes Polycom devices offline for some customers. (zoomstatus.com)Resolved in the past 24 hours:
GitHub: Copilot Code Review could not complete reviews on September 28 from about 21:16 to 22:08 UTC. GitHub rated it critical and reverted a change to fix it. Other GitHub services were not affected. (githubstatus.com)Cloudflare: elevated error rates for traffic in Eastern North America on September 28, from about 19:50 to 20:20 UTC. Also a short incident with stuck Flux hotswap models on Replicate, whose status page now lives on Cloudflare's status site.Other recent incidents from the past 72 hours, all resolved: Cloudflare R2 problems in Western North America and a Cloudflare CASB incident (September 26).
Consumer ISPs: crowd-sourced sites showed small, scattered report spikes overnight for Spectrum, Cox, Megacable and a few regional networks. Nothing points to a large outage at any of them. On September 28, user reports also spiked for Life360 and several online games (Roblox, GTA Online, Rainbow Six Siege, Escape from Tarkov), but none of these was confirmed as a major outage.
Routing: no significant new BGP incident was reported today.
Overall: the internet is up. Keep an eye on Akamai in Canada and on APAC connectivity while the subsea repairs continue.