On 18 November 2025, Cloudflare experienced a major outage. Many websites, apps, and services that rely on it suffered errors, slowdowns, or full unavailability — including X (formerly Twitter), ChatGPT, Spotify, Uber, and others.
The incident started being reported around 11:20 UTC (≈6:20 a.m. ET) when Cloudflare saw an “unusual traffic spike”. Cloudflare’s status page recorded the event as an “internal service degradation” impacting multiple services.
The root cause: a configuration file (used for bot-mitigation / threat traffic) grew too large, triggering cascading system crashes inside Cloudflare’s traffic-handling infrastructure.
The company deployed a fix and declared the incident resolved around 14:30 UTC (≈2.5 hours later), though monitoring continued.
Cloudflare reported that the sequence of events was as follows: a routine configuration change led to an automatically generated file expanding beyond the expected size, which triggered errors in the traffic-handling system, resulting in service degradation.
Many users experienced 500-error pages (“Internal Server Error”), failed loads, or no access to services that depended on Cloudflare. Even the outage-tracking site Downdetector was impacted, since it uses some of the same infrastructure.
Popular websites like OpenAI’s ChatGPT experienced outages due to Cloudflare issues. Source: Reuters
This event will contribute to the ongoing discussion regarding the extent to which a limited number of infrastructure providers support the internet and the associated risks that come with it.