Internet & cloud outages today

The internet is broadly healthy today. AWS, Google Cloud, Azure, Azure DevOps, Cloudflare, GitHub, OpenAI, Meta, Stripe, Vercel, Heroku, Discord and Atlassian all report normal operations. There is no sign of an internet-wide event — no significant BGP leak, no submarine cable break, no DNS failure. ACTIVE: ELEVENLABS MEDIA DELIVERY ElevenLabs is investigating media that will not load across Studio, Audiobooks, Image & Video, Music and Dubbing. Both its ElevenAPI and ElevenCreative components are marked degraded. The incident opened at 08:49 UTC and, as of this report, has neither a cause nor a resolution. Source: https://status.elevenlabs.io/ ACTIVE, NARROW SCOPE • Twilio has four open issues — three carrier-specific SMS delivery delays (ICE Costa Rica, Dukagjini Kosovo, Telefónica El Salvador) and one where recently upgraded accounts are still hitting trial-account restrictions. Twilio's Carrier Network and Developer Tools sections show degraded performance; the core APIs do not. Source: https://status.twilio.com/ • Slack has carried a minor incident since August 13: workspaces with custom data retention policies have trouble reaching historical messages. Messaging and workspace administration are flagged; login, files, huddles, search and apps are not. Source: https://slack-status.com/ • Akamai's edge delivery problems in India, open since August 22, are in monitoring after a fix. Akamai has ruled out the third-party provider it initially suspected. Source: https://www.akamaistatus.com/ • DigitalOcean is running extended Spaces maintenance in London, Singapore and Sydney, with brief interruptions possible. Planned work, not an outage. THE DAY'S CROSS-PROVIDER STORY: AI MODEL ERRORS Overnight, Anthropic logged elevated errors on requests to Claude models including Opus 5 and Fable 5, from 9:50pm to 12:36am PT. The same window showed up downstream at two other vendors: GitHub reported reduced Copilot availability for the Fable model "due to an upstream provider," and Cursor reported elevated Claude Opus 5 errors from 05:38 to 08:51 UTC. All three are resolved. It is a useful reminder that AI coding tools now share a small number of model backends, so one provider's bad hour becomes several products' bad hour. RESOLVED IN THE LAST 72 HOURS • Cloudflare: Durable Objects and D1 degraded in Hong Kong and Singapore, and elevated HTTP 5xx errors in Chicago. Both minor, both resolved earlier today. Maintenance continues in Ashburn, Montréal and Astana. • Zoom: roughly 40 minutes of degraded meeting chat today, resolved. • GitHub: on August 21, SSH Git operations fell below 95% success for about four minutes, and Copilot agent task creation failed intermittently. Git over HTTPS was unaffected. • DigitalOcean: a significant interruption in the MKC1 region on August 19 affecting GPU Droplets is largely restored, with GPU node recovery continuing. ONE SOURCING CAVEAT AWS's public RSS feed is currently serving stale March 2026 entries, so today's AWS read comes from the Health Dashboard and secondary reporting rather than the feed. Neither showed an open broad event. Third-party monitors noted a short AWS warning window on August 23 that we could not corroborate against AWS's own dashboard, so we are not treating it as a confirmed incident. Overall: incidents detected, none widespread. The internet is up.

The internet is operating normally today. We found no widespread outage, routing failure, or DNS disruption across the providers we monitor. A handful of narrower incidents are worth noting. Akamai — Edge Delivery issues in India. Akamai began investigating around 11:40 UTC and attributed the disruption to a third-party service provider. A fix has been implemented and the incident is now in a monitoring state, with service resuming normal operation. Akamai is separately reporting degraded performance on its Data Services and Reporting components. This was the day's most substantive infrastructure item, but the impact was regional rather than global. (akamaistatus.com) Slack — an incident open since August 13 continues to affect access to historical messages for workspaces that have custom data retention policies enabled. Messaging and workspace/org administration are flagged; login and SSO, connectivity, files, huddles, search, and workflows are normal. Narrow in scope, but still unresolved after nine days. (slack-status.com) Twilio — several country-specific SMS delivery problems remain open, including delivery delays to Telefonica El Salvador, delivery failures to Dukagjini in Kosovo, and delayed delivery receipts to ICE in Costa Rica. A scheduled London (IE1) maintenance window affecting Voice, Messaging, Authentication, and Interconnect also ran through the day. Earlier SMS issues in Bulgaria, Oman, and Vietnam were resolved. (status.twilio.com) Everything else we track reports normal operation: Cloudflare, Google Cloud, Microsoft Azure, GitHub, OpenAI, Anthropic, Discord, Atlassian, Zoom, Vercel, Heroku, DigitalOcean, Stripe, and the AI platforms on our list (Cursor, Groq, Replicate, ElevenLabs, Cohere, Stability AI). Cloudflare resolved a minor network connectivity issue in Dubai (DXB) at 02:57 UTC. One caveat on crowd-sourced signal: third-party report trackers showed small spikes for Cloudflare and X during the afternoon. Neither company confirmed an incident, and we found no corroborating evidence, so we are treating those as unverified user-report noise rather than real events. Recently resolved, past 72 hours: DigitalOcean restored Container Registry and Spaces access on August 21 after customers were unable to push or pull images, and recovered its MKC1 region on August 19 following an interruption to GPU Droplets, Kubernetes nodes, and serverless inference. Zoom resolved a brief Canvas outage in its US region on August 21 and a file-search problem in Zoom Hub on August 22. GitHub resolved intermittent failures creating Copilot Cloud Agent tasks on August 21. A note on sourcing: the AWS Health Dashboard renders through JavaScript and did not return readable content to us this run. We found no news reports or corroborating signals pointing to an AWS incident, and the aggregator flags we did see for AWS look like stale component noise rather than a live event — but our AWS read today is indirect rather than directly confirmed. Overall: incidents detected, none internet-wide.

No internet-wide disruption today. The major cloud, CDN, and platform providers are largely operating normally, with a handful of service-specific incidents open and no new BGP routing or submarine cable faults detected. Active incidents Akamai has Data Services and Reporting degraded, with a data-availability issue affecting "Traffic by Hostname" reporting. Root cause was identified on August 20 at 18:32 UTC and mitigation remains in progress. Content delivery itself is unaffected. Source: https://www.akamaistatus.com/ Slack continues an incident open since August 13: channels in workspaces with custom data retention policies cannot serve historical messages. Retention settings have been restored, but the retention jobs themselves remain on hold. Source: https://slack-status.com/ Twilio is partially degraded, affecting Bulk Export, some carrier lookups, and Zipwhip toll-free SMS to North America. Two separate issues resolved today: Japan voice services and Altan Redes SMS in Mexico. Source: https://status.twilio.com/ Zoom reports degradation in its Meta integration for Zoom CX third-party clients, with the root cause identified. Source: https://www.zoomstatus.com/ Long-running conditions AWS me-south-1 (Bahrain) remains unavailable and me-central-1 (UAE) remains impaired, both dating to conflict-related infrastructure damage in late March 2026. AWS has indicated restoration is measured in months rather than days. This was verified against independent reporting rather than the status feed alone. Source: https://health.aws.amazon.com/health/status Resolved in the past 72 hours Google Cloud cleared a us-west1 incident on August 20 that ran 3 hours 40 minutes and touched more than 26 products. GitHub resolved a Copilot Cloud Agent degradation spanning August 20 and 21. Cloudflare closed a cluster of short regional incidents. Anthropic resolved model and connector errors on August 20. DigitalOcean restored Container Registry and Spaces, with GPU Droplets in MKC1 still recovering. AWS cleared a Connect metrics issue in ap-northeast-1. Operating normally Microsoft Azure, Azure DevOps, OpenAI, Discord, Atlassian, Vercel, Heroku, Stripe, Cursor, Groq, Replicate, ElevenLabs, Cohere, Stability AI, and Meta all report all-clear. Bottom line: the internet is up. Today's open issues are service-specific rather than internet-wide.

No widespread internet disruption today. The major cloud, CDN, and platform providers are largely operating normally, though a handful of regional and service-specific incidents are open. Active incidents Cloudflare is monitoring two minor regional incidents: increased errors for Durable Objects and downstream services in Indonesia (opened 13:49), and network performance issues in India (opened 12:01). Both are classified as minor and are in the monitoring stage. A planned maintenance window at the Chicago (ORD) facility is also in progress. Source: https://www.cloudflarestatus.com/ OpenAI is investigating errors affecting some users of Thinking in ChatGPT. The incident was opened very recently and the impact scope has not yet been stated. Source: https://status.openai.com/ Twilio reports SMS delivery delays from a subset of Twilio short codes to T-Mobile US subscribers. The SMS short code and toll-free components in North America are showing degraded performance. Source: https://status.twilio.com/ Slack has an open incident affecting access to historical messages in workspaces with custom data retention policies enabled, ongoing since August 13. All other Slack components are operational. Source: https://slack-status.com/ Akamai is continuing to monitor its Certificate Provisioning System after Control Center UI timeouts and API errors on August 19 between 03:12 and 03:34 UTC. Normal operations have resumed. Source: https://www.akamaistatus.com/ AWS Middle East (UAE), region me-central-1, remains impaired, and AWS continues to advise customers to run workloads in other regions. This is a long-running condition dating back to the March 1 physical damage to two of the region's three availability zones, not a new development today. All other AWS regions are operating normally. Source: https://health.aws.amazon.com/health/status Resolved in the past 72 hours Fastly resolved a global CDN performance impact spanning August 19 and 20 that affected multiple POPs, during which some purge requests were lost between 17:02 and 23:24 UTC. A separate reduced cache performance issue affecting Compute customers was also resolved earlier today. GitHub resolved a significant 7 hour 47 minute degradation on August 17 that caused elevated errors and latency across Issues, Pull Requests, APIs, Actions, and Copilot, peaking near 20 percent web and API error rates. GitHub attributed it to load balancer network saturation in its Central US datacenter combined with a sidecar pod that failed to scale. DigitalOcean cleared an August 19 issue in its MKC1 region affecting GPU Droplets. AWS resolved elevated EC2 instance launch errors in a single London (eu-west-2) availability zone on August 19. Cloudflare also resolved network performance degradation in Singapore and elevated error rates in its Chicago region. Operating normally Google Cloud, Microsoft Azure, GitHub, Discord, Atlassian, Vercel, Heroku, DigitalOcean, Zoom, Stripe, and X all report all-clear. Crowd-sourced outage report volumes show no unusual spikes, and we identified no significant BGP routing or DNS incidents. Bottom line: the internet is up. Today's open issues are regional or service-specific rather than internet-wide.

The internet is up. No widespread outages today — just a handful of minor, isolated incidents at individual providers. DigitalOcean is investigating degraded performance in its MKC1 region (https://status.digitalocean.com), affecting GPU workloads and some Kubernetes worker nodes across multiple racks since 11:10 UTC. The issue is contained to that single region. Slack is still working through an incident open since August 13: users in workspaces with custom data retention policies may be unable to load older messages (https://slack-status.com). Core messaging is unaffected. Twilio reports carrier-level SMS delivery delays or failures to a few networks in Belgium, Canada, and Brazil (https://status.twilio.com) — narrow in scope and typical day-to-day carrier churn. Recently resolved: GitHub had a major incident on August 17 lasting nearly 8 hours, with roughly 20% of web and API requests failing at peak due to network saturation on its Central US load balancers; it was fully resolved the same day and GitHub has been stable since (https://www.githubstatus.com). Cloudflare resolved two minor issues in the past day (Durable Objects errors in Hong Kong, elevated CDNJS errors), and Akamai closed out a multi-day Miami latency issue on August 18. Everything else checks out: AWS, Google Cloud, Azure, Cloudflare, Akamai, OpenAI, Discord, Atlassian, Zoom, Vercel, Heroku, and Stripe all report normal operations. No BGP, DNS, or routing anomalies detected today. Crowd-sourced outage trackers show only routine, scattered ISP reports.

Google's Home and Nest smart-home platform suffered a global outage today. Users across multiple regions reported Nest Hub displays and Google Nest/Home speakers failing to respond to voice commands, routines and automations not running, timers failing, and devices stuck on loading screens or returning 'Hmm, something went wrong.' Crowd-sourced reports pointed at Google's cloud backend rather than the devices themselves — roughly 74% of Downdetector complaints cited server connection problems. The incident does not appear on the Google Cloud status dashboard (consumer Home/Nest services aren't tracked there), and Google had not issued a statement at the time of writing, but user reports indicate service is recovering. Coverage: https://phandroid.com/2026/08/18/its-not-just-you-google-home-and-nest-suffers-global-outage-as-users-report-problems/ GitHub has two minor active incidents this morning: intermittent failures loading runner-group and runner-permission pages (monitoring) and an investigation into degraded performance for some services (https://www.githubstatus.com/). This follows yesterday's larger, now-resolved incident involving Copilot authentication failures and broader service degradation. Smaller ongoing items: Slack is still working through a narrow issue (since Aug 13) where workspaces with custom data-retention policies have trouble accessing historical messages (https://slack-status.com/). Akamai continues to implement a fix for connection drops in Miami, with no customer impact reported. Twilio shows routine regional SMS carrier degradations (UK, Brazil, French Caribbean). Spectrum shows only scattered, localized user reports — normal background noise, not a network event. Everything else checked is clear: AWS, Google Cloud, Azure, Cloudflare, Fastly, OpenAI, Discord, Atlassian, Vercel, Heroku, DigitalOcean, Stripe, Zoom, and Meta all report normal operations, and there are no signs of BGP, DNS, or undersea-cable problems. Core internet infrastructure is healthy.

No internet-wide disruption today. Core cloud, CDN, and platform providers are largely healthy, with a handful of regional and service-specific incidents worth knowing about. The most notable active issue is AWS Direct Connect at the Equinix FR5 facility in Frankfurt, Germany. A facility infrastructure problem (reported as a cooling outage) has caused packet loss and connectivity degradation for customers whose Direct Connect circuits terminate solely at that site since roughly August 15. AWS says customers with redundant connections at other locations are unaffected, and remediation is underway pending replacement network hardware. Details: https://health.aws.amazon.com/health/status Separately, AWS's Middle East (UAE) region, me-central-1, remains impaired — the long-running aftermath of physical damage to two availability zones on March 1. AWS continues to advise affected customers to run workloads in other regions, with full recovery expected to take months. This is ongoing background, not a new development. Elsewhere: Twilio is working through several regional issues, including elevated errors on Interconnect at its Frankfurt edge, SMS delivery delays to networks in Colombia and Brazil, voice call delays to Iraq, and elevated Microsoft 365 IP blocks affecting some SendGrid email senders (https://status.twilio.com). Slack has an active incident, ongoing since August 13, affecting access to historical messages for workspaces with custom data retention policies (https://slack-status.com). Akamai is implementing a fix for connection drops and latency in Miami, though it reports no current customer impact (https://www.akamaistatus.com). Resolved in the past 72 hours: Cloudflare closed out a minor Durable Objects error-rate incident this morning; GitHub had several brief degradations on August 12–13 (webhooks, pull requests, logins) that were all resolved; Discord fixed a call-setup issue on August 13; and DigitalOcean resolved a volume attach/detach issue in FRA1 on August 13. Google Cloud, Azure, Heroku, Vercel, Zoom, Atlassian, OpenAI, Discord, and DigitalOcean all report normal operations. A few status pages (Fastly, Stripe, Meta, X) could not be read during this check, but no outage signals for those services appeared in crowd-sourced reports or news searches.

The internet is operating normally today. No widespread outages, routing failures, or DNS problems are detected across the major cloud, CDN, and platform providers we monitor. Core infrastructure is healthy. Amazon Web Services, Google Cloud, Microsoft Azure, Cloudflare, Fastly, GitHub, OpenAI, Discord, Atlassian, and Vercel all report normal operations with no active incidents. A few minor, localized issues are worth noting but none affect the broader internet: - Akamai has an identified networking issue in its Miami region (elevated latency / connection drops) that it says customers should not currently see impact from. - Twilio shows degraded performance on some carrier routes and two narrow active incidents (SMS delivery delays to a Brazilian carrier, and elevated Microsoft 365 IP blocks affecting some SendGrid email delivery). - Slack has an ongoing incident affecting access to historical messages only for workspaces using custom data-retention policies. Recently resolved (past 72 hours): DigitalOcean cleared a Block Storage issue in its FRA1 (Frankfurt) region on August 13, and Discord resolved a brief call-establishment issue the same day. The multi-hour GitHub Actions outage from August 6–7 remains fully resolved. Bottom line: the internet is up and stable. Sources: cloudflarestatus.com, health.aws.amazon.com, status.cloud.google.com, azure.status.microsoft, githubstatus.com, akamaistatus.com, status.twilio.com, slack-status.com.