Othercloudflare outage
Summary (tl;dr)
A major global Cloudflare outage on November 18, 2025, disrupted access to numerous popular websites and online services, including X (formerly Twitter) and ChatGPT, causing widespread "500 Internal Server" errors and a confusing "Please unblock challenges.cloudflare.com to proceed" message.
Essential Background
Cloudflare is a vital internet infrastructure company based in the USA, providing critical services like content delivery networks (CDN), DDoS protection, DNS, and security to millions of websites globally. It acts as a "gatekeeper" that monitors web traffic to protect sites from malicious attacks and verifies users as human, often through automated challenges. Due to its extensive reach, any significant issue with Cloudflare's network can lead to widespread internet disruptions.
The Full Story
On November 18, 2025, Cloudflare experienced a significant service degradation, which the company acknowledged as its "worst outage since 2019". The incident, which commenced around 11:20 UTC and was largely resolved by 14:30 UTC, stemmed from a latent bug in the generation logic of a Bot Management feature file. This bug caused a critical configuration file to grow beyond its expected size, leading to a crash in the software system responsible for handling traffic for several Cloudflare services.
As a result, internet users worldwide attempting to access websites reliant on Cloudflare's infrastructure, such as X, ChatGPT, Canva, and Perplexity AI, encountered widespread "500 Internal Server Errors". Many users also reported seeing a perplexing message: "Please unblock challenges.cloudflare.com to proceed." While this message typically appears when a user's browser or security settings block Cloudflare's human verification tool (Turnstile), in this instance, it was a misleading consequence of Cloudflare's own internal server failures.
Why It Matters
This global Cloudflare outage highlights the profound interdependence of modern internet services on a few foundational infrastructure providers. The widespread disruption, affecting major platforms across social media, artificial intelligence, design, and gaming, underscored how a single point of failure can have far-reaching implications for global online accessibility and user experience. Such incidents prompt critical discussions about the internet's resilience and the ongoing need for diversified infrastructure and robust contingency planning to mitigate the impact of future system failures.
Geographic Location
- London, England, United Kingdom (Cloudflare disabled an encryption service called Warp as part of remediation efforts)