On November 18, 2025, major internet services around the world experienced simultaneous outages, causing widespread user inconvenience. Social media platforms like X and internet infrastructure provider Cloudflare experienced outages one after another, with thousands of users reporting connectivity issues.
According to Downdetector, over 5,600 users reported X connection errors this morning. Messages indicating that posts could not be loaded appeared on the screen, and in some areas, service was completely down. CloudFlare also confirmed the outage, which is affecting multiple customers, and said it is working to identify the cause and restore the service.
Cloudflare is known as a core infrastructure service that powers a significant portion of websites worldwide. It handles massive traffic distribution and security functions, serving as the hidden infrastructure of the Internet. Around 11:37 AM, multiple services relying on Cloudflare experienced simultaneous errors, with some sites displaying an internal server error code of 500. Cloudflare's own services also exhibited instability, with the Cloudflare status page briefly displaying broken styles.
The impact of the outage spread quickly. Platforms including OpenAI, Facebook, AWS, bet365, Canva, Spotify, and League of Legends reported issues. Online game access was interrupted or unstable, and some payment services experienced delays in processing orders. Downdetector, which relies on Cloudflare, was unable to properly report the outage for a while.
Graeme Stewart, Check Point's head of public affairs, explained that this outage exposed the vulnerability of large-scale platform dependencies. He analyzed that a single disruption in a massive infrastructure service can simultaneously impact multiple organizations. He pointed out that many companies maintain a structure that relies on a single path for core functions, leaving no alternative path in the event of an outage.
Cloudflare stated that it is investigating whether the outage was related to scheduled maintenance at its Santiago data center and that some functionality is being restored. However, it maintains a warning that the error rate may be higher than usual.
