Cloudflare Blames “Unusual Traffic Surge” for Global Outage Hitting X and AI

Cloudflare has officially broken its silence regarding the massive global outage that crippled a significant portion of the internet on Tuesday, November 18, 2025. The infrastructure giant, often described as the web’s “hidden gatekeeper,” confirmed that a sudden and unexplained spike in traffic triggered the cascade of failures that took down major platforms including X (formerly Twitter) and OpenAI’s ChatGPT.
In a statement reported by The Guardian, a Cloudflare spokesperson revealed the timeline of the disruption. “We observed a surge in unusual traffic to one of Cloudflare’s services, which began at 11:20 AM,” the spokesperson said. “This led to errors in the operation of some traffic passing through the Cloudflare network.” While the company noted that most traffic continued to flow normally, the specific services affected were critical enough to cause widespread “500 Internal Server” errors for millions of users worldwide.
The company’s status updates painted a picture of engineers scrambling to contain the damage. Initial reports indicated a “higher than usual error rate” as troubleshooting began. By 12:21 GMT, Cloudflare announced that a fix was being implemented and that services like Cloudflare Access and the WARP encryption tool were recovering to pre-incident levels. “We have made changes that have allowed Cloudflare Access and WARP to resume operation,” the company stated, adding that they were still working to fully restore services for application clients.
The root cause of the traffic spike remains a mystery. “We do not yet know the cause of the surge in unusual traffic,” the spokesperson admitted. “We are fully focused on ensuring that all traffic is served without errors. After that, we will turn our attention to investigating the cause.”
Speculation has arisen regarding scheduled maintenance that was planned for the same day in data centers across Los Angeles, Atlanta, Santiago, and Tahiti. However, it remains unclear if these routine operations were linked to the sudden collapse. In a desperate bid to stabilize the network during the crisis, engineers temporarily disabled the WARP service in London, warning users of immediate connection issues before eventually re-enabling access.
Alan Woodward, a cybersecurity expert at the Surrey Centre for Cyber Security, highlighted the immense scale of the incident. He described Cloudflare as the “biggest company you’ve never heard of,” emphasizing its critical role in verifying human users and shielding websites from DDoS attacks. When a central node like Cloudflare fails, the ripple effects are immediate and global. This incident is particularly alarming as it comes less than a month after a similar outage at Amazon Web Services, raising serious questions about the resilience of the centralized internet infrastructure.