No internet-wide outage was detected today. Core infrastructure is healthy, but a few provider-level incidents are active.
Active incidents
Anthropic (Claude): As of about 09:14 UTC, the status page lists elevated errors for Claude Sonnet 5.5 (status: Investigating). claude.ai, the Claude API, Claude Code and Claude Cowork show partial outages; the Console is operational. Similar elevated-error incidents were resolved on Oct 24 and Oct 25. Source: status.claude.comTwilio: Several carrier-specific messaging and voice issues are open, including voice call failures to networks in Germany (investigating), SMS failures in Brazil and Mauritania, voice failures in Romania (fixes in progress), and long-running delays in Thailand and for GCI MMS in the US. Programmable Messaging is listed as degraded. These are regional/carrier-level, not global. Source: status.twilio.comOpenAI: One incident is being monitored after a mitigation, with elevated error rates for hosted Code Reviews and Security Reviews on private GitHub repositories. Source: status.openai.comAkamai: Edge delivery issues between Indonesia and Singapore were fixed on Oct 25 and are still being monitored. Akamai is also assessing two sets of vendor vulnerability advisories (Brotli and F5/NGINX) and reported no evidence of customer impact so far. Source: akamaistatus.comRecently resolved
GitHub: elevated latency for REST API, Actions, Issues and Pull Requests on Oct 26 (resolved 18:57 UTC); no incidents reported today so far. Source: githubstatus.comAll clear at time of checking
Cloudflare (scheduled data center maintenance only), AWS, Google Cloud, Azure, Discord, Slack, Atlassian, Vercel, DigitalOcean, Groq, Cohere, ElevenLabs and Stability AI reported no active incidents. Zoom lists only planned maintenance. Heroku, Stripe, Azure DevOps and Meta pages returned no readable incident data, so their status is unconfirmed. Replicate and Cursor status pages redirected elsewhere and were not checked. Web searches and BGP searches surfaced no confirmed current event. Downdetector was not accessed directly.
Overall: incidents detected, none widespread.