Cloudflare’s global outage takes down sites in the US and around the world, including X, ChatGPT and Spotify this Tuesday
Cloudflare registered an instability in its global network around 11am this Tuesday, November 18, 2025, affecting access to multiple websites and services

Cloudflare registered an instability in its global network around 11am this Tuesday, November 18, 2025, affecting access to multiple websites and services on Estados Unidos and in other countries. Plataformas like X, ChatGPT and Spotify displayed connection errors for users across the world, due to internal degradation in the company’s services. The company identified an unusual traffic spike as an initial trigger and initiated fixes to mitigate the issue.
The incident occurred during scheduled maintenance at data centers in the US, such as at Los Angeles and Atlanta, although the direct relationship has not been confirmed. Usuários reported 500 error messages and failed loads, impacting daily online content traffic.

American platforms impacted
X faced total interruptions in feeds and posts, with millions of accesses blocked on Estados Unidos. OpenAI, responsible for ChatGPT, recorded blocks in artificial intelligence queries for global users.
Discord, widely used for in-game communications, has stopped responding on US servers. Serviços and Spotify exhibited slow audio streams during the peak of the failure.
These effects highlight the dependence on centralized infrastructures for day-to-day operations.
Global Services Affected
The flaw spread to international platforms, including Facebook, which saw errors in logins and page loads. Canva, a collaborative design tool, has become inaccessible for team editing around the world.
Online games like League of Legends have recorded disconnections in active matches. E-commerce and digital banking Sites experienced delays in transactions, with integrations via Cloudflare’s API temporarily failing.
- Facebook: Erros in feeds and messages for 30 minutes;
- Canva: Bloqueio on file uploads;
- League of Legends: Desconexões on global servers.
Answers from Cloudflare
Company engineers issued updates confirming partial recovery in services such as Access and WARP. Taxas error has returned to normal levels in regions like Londres and parts of the US.
The company worked with third-party providers to isolate the problem, prioritizing restoration of APIs and dashboards. Até at noon, 80% of services operated without significant interruptions.
Maintenance in data centers at América of Sul and Norte coincided with the event, but were completed without extending the outage.
Continuous monitoring prevented further spread to corporate customers.
History of similar outages
Past events reveal patterns in global CDN failures. In October 2025, an AWS outage affected similar services, including Snapchat and Reddit, for hours.
Cloudflare reported incidents in 2024 related to software updates that impacted partner networks. Esses cases involve traffic spikes that overload centralized routing, affecting up to 10% of global web traffic.
Experts note that the company processes 81 million HTTP requests per second on average, expanding the scope of any degradation. In previous outages, millions of users in dozens of countries reported impacts, similar to today, with durations ranging from 20 minutes to several hours depending on the root cause.
Measures for users and companies
It is recommended to use independent monitoring tools to track network status during events like this. Empresas can diversify CDN providers to reduce single dependency risks.
Individual users opt for connections via VPNs or offline caches to access essential content. Affected Plataformas activated secondary routes to minimize operational losses.
- Monitor via apps like DownDetector, when available;
- Activate offline modes in streaming apps;
- Contact service support for real-time updates.
Full normalization proceeded in stages, focusing on high-traffic regions in the US and Europa.
US user reports
Professionals in Nova York and Los Angeles shared difficulties in accessing work tools in the morning. Um developer Canva mentioned migration to local editors during the outage.
On São Francisco, League of Legends players reported pauses in online sessions. Usuários from X in Estados Unidos moved to direct messaging in alternative apps to maintain communications.
These reports emphasize the disruption to professional and entertainment routines dependent on digital services.






