Internet & cloud outages — April 2026

The internet is operating normally today with only isolated incidents detected. Cloudflare is conducting scheduled maintenance at its Minneapolis (MSP) datacenter from 07:00–16:00 UTC, and is monitoring previously delayed Cloudflare Access audit logs that are now clearing. OpenAI reports a partial disruption affecting write actions for some ChatGPT Workspace Connectors; workspace administrators can manually restore permissions via the connector settings page until OpenAI fully resolves the issue. The major cloud platforms — AWS, Microsoft Azure, and Google Cloud — report no broad incidents. GitHub is operating normally. Regional reports include intermittent Sky Broadband connectivity issues in the UK. Recently resolved incidents (past 72 hours): • Apr 28 — Brief AWS multi-services operational issue • Apr 25 — Zayo Group (Tier 1 carrier) multi-region disruption • Apr 22 — Verizon Business multi-region outage (~63 minutes) Overall, core internet infrastructure remains stable and no widespread or systemic disruptions are in progress. Sources: Cloudflare Status, Azure Status, Google Cloud Service Health, OpenAI Status, AWS Health Dashboard.

The internet is broadly operational on April 28, 2026, with no widespread outages reported across Tier-1 cloud providers or core infrastructure. A small number of localized incidents are being tracked. Active or recently monitored: - Cloudflare is monitoring network performance in its Istanbul (IST) region after implementing a fix earlier today (identified 08:45 UTC, fix in place by 09:21 UTC). All major Cloudflare services remain operational globally. (https://www.cloudflarestatus.com) - DownDetector showed brief, small overnight reporting spikes for several services including T-Mobile, Quantum Fiber, Sky Broadband, Microsoft 365 / Outlook, and GitHub. None correlate to a broader infrastructure-level event. Recently resolved (past 72 hours): - AWS reported regional connectivity issues affecting the Paris region (eu-west-3) on April 27, with impact to API operations and several services including AWS Lambda and AWS Config. The event has been resolved. - Azure resolved an East US provisioning and scaling issue from April 24–25 attributed to a recently deployed update. - Cloudflare resolved network performance degradation in Brazil and a Turnstile challenge-solve issue on April 27. Status pages reporting all clear right now: Microsoft Azure, Google Cloud, OpenAI. The internet is up.

The internet is operating normally on April 27, 2026. Major cloud, CDN, and platform providers — including AWS, Microsoft Azure, Google Cloud, Cloudflare, Fastly, and OpenAI — are reporting no active widespread incidents. Cloudflare resolved two minor issues earlier today: failed Turnstile challenge attempts (resolved 07:09 UTC) and a brief network performance degradation affecting customers in Brazil (resolved 04:50 UTC). A routine scheduled maintenance window is currently in progress at Cloudflare’s Salt Lake City (SLC) datacenter, which may cause slight latency increases for traffic routed through that location. No significant BGP routing incidents, DNS failures, or undersea cable disruptions have surfaced in the past 24 hours. Crowd-sourced outage trackers show no unusual spikes against major consumer services. The internet is up and operating normally.

The broader internet is operating normally today, with one notable regional incident on Microsoft Azure. Microsoft Azure - East US (Active) Azure is investigating a partial outage in its East US region that began around 11:39 UTC on April 24, 2026. Affected services include Virtual Machines and VM Scale Sets, Azure Kubernetes Service (AKS), Service Fabric, Application Gateway, Azure Virtual Desktop, Azure Databricks, Azure Data Explorer, Azure Cache for Redis, Azure Synapse, and Azure Database for PostgreSQL flexible servers, among others. Microsoft has identified a suspected triggering deployment and a rolling rollback is underway across affected availability zones. Customers in East US may see failures or delays when provisioning or scaling resources, along with intermittent connectivity for existing workloads. Microsoft recommends deploying new workloads to alternate regions and pausing CI/CD pipelines targeting East US until the issue is mitigated. Source: https://azure.status.microsoft Cloudflare Cloudflare worked through a handful of small issues earlier today, including intermittent HTTP 5xx errors on Cloudflare Access authentication (fix implemented, monitoring), brief Healthchecks monitoring and alerting disruption, a Load Balancing dashboard display bug (CDN unaffected), WARP connectivity issues in Bulgaria, and a short network performance degradation in Miami. All of these have been resolved or are under monitoring with limited user impact. Scheduled maintenance is being performed today at the Seattle (SEA) and Montreal (YUL) datacenters, which may cause minor regional latency variations. Source: https://www.cloudflarestatus.com Other Major Providers AWS, Google Cloud, GitHub, and OpenAI all report normal operations with no broad incidents. No significant BGP hijacks, DNS failures, or undersea cable issues have been reported today. Bottom Line Outside of the Azure East US regional incident, the internet is operating normally. Users and operators outside that region are unlikely to see any impact.

The internet is operating normally today. Major cloud providers — AWS, Azure, and Google Cloud — all report no active severe incidents, and OpenAI is fully operational. A few minor issues worth noting: - Cloudflare is investigating an issue with its Dashboard API that is affecting the loading of R2 buckets and causing elevated API latency (investigating as of 10:37 UTC). Cloudflare is also running previously scheduled maintenance at its Seattle (SEA), Buenos Aires (EZE), and Minneapolis (MSP) datacenters, which may cause brief latency increases for traffic routed through those regions. - Grok (xAI) users have reported intermittent errors and response delays on the web, mobile apps, and X platform integration since Wednesday evening. Official status pages show services as operational; the disruption appears load-related rather than an infrastructure failure. No BGP routing incidents, DNS disruptions, or undersea cable issues have been reported in the past 24 hours. Downdetector activity across major consumer services is within normal ranges. Overall status: incidents detected, minor in scope. The internet is up.

The core internet is operating normally today. Major cloud platforms including AWS, Azure, Google Cloud, and Cloudflare report no widespread outages. Cloudflare has scheduled maintenance underway at its Seattle (SEA) and Salt Lake City (SLC) datacenters, which may cause minor latency increases in those regions but is not a service outage. Several platform-level incidents are being tracked: - OpenAI is investigating elevated conversation error rates for ChatGPT Enterprise, Business, and Education workspaces; a mitigation has been applied and recovery is being monitored. - Grok (xAI) is showing elevated user reports of accessibility issues starting around 00:00 ET. - Fastly is tracking elevated errors for its Fastly Application component, ongoing since April 21. - AWS has minor, regionally-scoped reports affecting UAE Cloud Map and ElastiCache, plus intermittent Management Console issues. Recently resolved over the past 72 hours: - April 20: ChatGPT global outage affecting 13,000+ users. - April 20: Fastly Ashburn (IAD) POP node crashes causing 5xx errors. - April 20: Cloudflare Johannesburg (JNB) datacenter performance issue. No BGP route leaks, DNS infrastructure problems, or undersea cable disruptions are currently reported. The internet is up.

The global internet is broadly operational today. No major cross-provider outages are in progress, though a handful of smaller incidents and scheduled maintenance windows are being tracked. Cloudflare is running scheduled maintenance at its Seattle (SEA) and St. Louis (STL) data centers through mid-day UTC, with brief latency increases possible as traffic reroutes. Cloudflare Sites and Zero Trust are currently showing degraded performance, and Cloudflare WARP is in monitoring after a connectivity issue that ended at 06:45 UTC on April 20. See https://www.cloudflarestatus.com for live updates. AWS is showing a limited, regional operational issue affecting AWS Cloud Map, Amazon ElastiCache, and the AWS Management Console in the UAE region; the rest of AWS is reported operational. Google Cloud, Microsoft Azure, GitHub, and OpenAI all report no active incidents at this time. Recently resolved (past 72 hours): AWS experienced a significant multi-service operational issue centered on a DNS failure on April 18 that briefly disrupted traffic for a number of global services. AWS has since restored normal operations. Overall: the internet is up, with minor localized incidents worth monitoring but no widespread disruption.

Global internet infrastructure is broadly stable on Monday, April 20, 2026, with no widespread outages or routing-level incidents detected. The major cloud providers — AWS, Microsoft Azure, and Google Cloud — all report normal operations, and no significant BGP or undersea cable events are active. A handful of low-impact service-level items are being tracked. Cloudflare has several items in a monitoring state: a fix has been applied for a ruleset rate-limiting accuracy issue affecting some Workers requests; an earlier WARP and Zero Trust connectivity problem was resolved at 06:45 UTC; and roughly 0.05% of customers may see HTTP analytics delayed by up to about six hours in the Cloudflare Dashboard. Cloudflare traffic and security services themselves remain unaffected. Scheduled maintenance at the Moscow (DME) datacenter is also in progress. OpenAI is monitoring a minor issue in which ChatGPT Business users may encounter errors after upgrading or adding new seats. GitHub, Fastly, and the major cloud status pages show no active incidents. Crowd-sourced outage trackers are not showing spikes that would suggest a widespread consumer-facing problem. Overall: the internet is up, with only isolated, low-severity items under watch. Sources: cloudflarestatus.com, status.openai.com, status.cloud.google.com, azure.status.microsoft, githubstatus.com, istheservicedown.com.

The global internet is broadly healthy today, with no major provider or backbone outages reported. Cloudflare is tracking two minor items: an Analytics API delay that has entered the monitoring phase after a fix was applied, and scheduled maintenance in the Sydney (SYD) datacenter (15:00–21:00 UTC) which may cause slight latency shifts for nearby users. Earlier scheduled maintenance in Frankfurt and Chicago has completed or is completing on schedule. AWS, Microsoft Azure, Google Cloud, GitHub, and OpenAI are reporting no active incidents on their public status dashboards. On the application side, a short burst of Bluesky error reports was noted in the early morning hours (peaking around 06:00 local), and smaller user-report spikes appeared for Vodafone AU and Fortnite. None reached the scale of a widespread outage and services appear to have recovered. No significant BGP leaks, DNS failures, or submarine cable incidents have been reported in the past 24 hours. Bottom line: the internet is up and operating normally, with a handful of isolated, low-severity incidents.

Most of the internet is running smoothly today. AWS, Google Cloud, Azure, OpenAI, and other major platforms report no active incidents. The main item of note is Cloudflare, which is experiencing a minor service outage with partial disruptions at edge locations across multiple continents. Affected locations include cities in Africa (Dar es Salaam, Djibouti City, Harare, Kigali, Kinshasa, and others), Asia (Bandar Seri Begawan, Malé, Naha, Kochi, Astana), Europe (Santa Venera, Skopje, Ljubljana), Latin America (multiple Brazilian cities, plus Bogotá, Arica, and Neuquén), and North America (Minneapolis, Norfolk). Traffic at these locations may be rerouted, causing slightly elevated latency for some users. Cloudflare also has scheduled maintenance windows today in Houston (15:30–23:00 UTC) and Frankfurt (23:30–06:00 UTC). Localized ISP outages have been reported in Tallahassee, Florida and Kansas City, Kansas, but these are regional in scope and not indicative of broader issues. Looking back at recent days, the Microsoft 365 and Azure outage that began on April 7 — which affected Teams, Outlook, and other services for users in the US and UK — has been fully resolved. Microsoft attributed the disruption to an Azure configuration change and confirmed services were restored after rerouting traffic through alternate infrastructure. No BGP incidents, DNS failures, or undersea cable issues have been reported today. Overall, the internet is up and functioning normally for the vast majority of users.

Overall, the internet is functioning normally today with one notable incident. OpenAI API is experiencing elevated 401 authentication errors across its API endpoints. The issue has been ongoing for approximately 16 hours as of 1:38 PM UTC. OpenAI has identified the problem and is working on implementing a mitigation. Users relying on OpenAI API integrations may encounter authorization failures. Other OpenAI services including Codex and Sora appear unaffected. Cloudflare is conducting scheduled maintenance at its St. Louis (STL) data center through 15:00 UTC today, which may cause slight latency increases for users routed through that location. Additional maintenance windows are planned for Santiago, Amsterdam, Osaka, and Frankfurt over the coming days. Some Cloudflare edge locations in smaller markets across Africa, Asia, Europe, and Latin America are reporting partial outages, which is typical for edge network operations. All major cloud providers are reporting normal operations: - Google Cloud: No incidents, all services operational - Microsoft Azure: All services operational, no active events - AWS: No widespread issues reported No BGP routing incidents, DNS failures, or undersea cable issues have been reported today. Recent resolved incident: On April 2, Microsoft 365 experienced a brief global disruption affecting access to office.com and the admin center, lasting approximately 74 minutes before full restoration. The internet is up.

The internet is largely healthy today with all major cloud providers reporting normal operations. Google Cloud, Microsoft Azure, and OpenAI are fully operational with no active incidents. Cloudflare is reporting a minor service outage affecting edge locations across multiple regions, including points of presence in Africa (Dar es Salaam, Djibouti, Kigali, Kinshasa), Asia (Bandar Seri Begawan, Malé, Naha), Europe (Malta, Skopje, Ljubljana), Latin America (several Brazilian cities, Bogotá), the Middle East (Basra, Najaf), and two US locations (Norfolk VA, Sacramento CA). Cloudflare is also performing scheduled maintenance at its Amsterdam datacenter. These edge-location disruptions may cause slightly elevated latency for users routed through affected PoPs but do not represent a core infrastructure failure. Several localized ISP outages were reported across the US, including connectivity issues in Idaho Falls (affecting schools), a fiber-optic cable break impacting Fremont Unified School District in California, and scattered reports from Chicago and Houston. These are all local provider issues, not internet-wide. No BGP routing incidents were detected today. Recently resolved: a Microsoft 365 global disruption on April 2 (lasted ~1 hour 14 minutes) and a brief iCloud Mail outage resolved by April 6. Overall, core internet infrastructure is stable. The Cloudflare edge issues are minor and localized to smaller points of presence.

No significant internet-wide outages or disruptions detected today. All major cloud providers — AWS, Google Cloud, and Microsoft Azure — are reporting normal operations with no active incidents. OpenAI services are fully operational across all components. Cloudflare is experiencing minor degraded performance at some edge locations across Africa, Asia, Latin America, and a few other regions, alongside scheduled maintenance at approximately 20 datacenters worldwide. These are routine operations and are not causing meaningful service disruption for end users. No BGP routing incidents, DNS failures, or undersea cable issues have been reported today. Looking back at the past 72 hours: a Spectrum/Charter broadband outage on April 2 affected thousands of customers across a dozen U.S. metro areas, and Microsoft 365 experienced a brief global disruption the same day due to degraded infrastructure in the Central U.S. datacenter (resolved within ~75 minutes). An iCloud Mail outage was also resolved by April 6. All of these are now fully restored. The internet is up and running normally.

Several notable but contained incidents are affecting parts of the internet today, though core infrastructure remains stable. **Cloudflare Dashboard & API Issues** Cloudflare is experiencing problems with its dashboard and API, with requests intermittently failing. A fix has been identified and is being implemented. CDN caching and core traffic routing remain unaffected. Separately, their Debian mirror service (cloudflaremirrors.com) is returning errors, and scheduled maintenance is underway at the Chicago (ORD) data center (06:00-12:00 UTC), which may cause slight latency increases for traffic routed through that location. **OpenAI ChatGPT Partial Outages** OpenAI reports that some users, particularly those on VPNs, are experiencing issues accessing ChatGPT. Twelve ChatGPT services are showing partial outages, including Search, GPTs, Deep Research, Voice mode, and Image Generation. Mitigation has been applied and the team is monitoring recovery. The OpenAI API itself remains largely unaffected at 99.98% uptime. **X (Twitter) Intermittent Issues** Users continue to report problems with the X mobile app and news feed, with reports of slow performance, inability to refresh timelines, and occasional error messages. These issues have persisted over the past 24 hours across multiple regions. **Major Cloud Providers** AWS, Google Cloud, and Microsoft Azure are all reporting normal operations with no active incidents. **BGP & Routing** No BGP hijacks, route leaks, or routing incidents have been reported today. **Local Outage** Socorro, Texas experienced a citywide phone and internet outage on April 6, which has been under investigation by local providers. Overall, the internet is functioning normally at the infrastructure level. The incidents today are platform-specific and do not indicate any broader connectivity issues.