A massive internet outage originating from infrastructure giant Cloudflare caused service failure for countless users and some of the world’s most popular digital platforms this morning, Tuesday, November 18, 2025.
Services including ChatGPT, the social media platform X (formerly Twitter), the gaming service League of Legends, and even firms like Amazon and Spotify were suddenly rendered inaccessible for a significant period.
I can confirm the immediate disruption was linked to Cloudflare’s core network.
The root cause of the crippling global disruption has been traced back to a seemingly routine operation. Cloudflare disclosed that a misconfiguration occurred during scheduled maintenance at its Santiago (SCL) data centre.
This maintenance was planned for a three-hour window between 12:00 and 15:00 UTC today. Instead of a smooth update, the error triggered a dangerous cascade of failures that propagated across Cloudflare’s entire worldwide network, the digital backbone that thousands of websites rely on for performance and security.
Users across the globe were immediately confronted with widespread HTTP 500 errors, system failures on Cloudflare’s own Dashboard, and non-functional APIs. For many, this meant complete inaccessibility of their chosen services for over an hour.
To contain the uncontrolled spread of the outage issue and attempt remediation, Cloudflare was forced to temporarily deactivate certain services for users in the United Kingdom.
Specifically, the company issued a notice stating: “During our attempts to remediate, we have disabled WARP access in London. Users in London trying to access the Internet via WARP will see a failure to connect.”
Cloudflare’s WARP is a zero-trust access tool, part of the company’s initiative to provide content delivery network (CDN) services and protection against attacks like DDoS. I believe this temporary shutdown was a last-ditch effort to stabilise their core network functions.
By 8:15 a.m. UTC, the company reported initial success in its recovery operations. Importantly, it confirmed that they had reinstated services for UK users, stating: “We have re-enabled WARP access in London.”
The company also noted that error levels for their Access and WARP services had “returned to pre-incident rates.” While this shows a partial return to normal, the company is still investigating the full extent and complete root cause of the internal service degradation.
