San Francisco – December 6, 2025 – A widespread internet disruption on Friday morning briefly knocked offline major platforms including President Donald Trump’s Truth Social, popular design tool Canva, and even the outage-monitoring service Downdetector itself, in the latest reminder of how dependent the modern web has become on a small handful of infrastructure providers.
The outage began around 8:47 UTC (3:47 a.m. Eastern Time) and lasted approximately 25–30 minutes. Users trying to access affected sites were met with generic Cloudflare error pages, most commonly “Error 1016: Origin DNS error” or “Error 503: Service Unavailable,” along with a unique Cloudflare Ray ID. Reports flooded in from North America, Europe, Australia, and the Middle East almost simultaneously.
Cloudflare, the San Francisco-based company that provides security, performance, and traffic-routing services to roughly one-fifth of all websites worldwide, quickly acknowledged the problem. At 9:20 UTC, Chief Technology Officer Dane Knecht posted on X: “We are aware of the issue impacting the availability of Cloudflare’s network… This was not an attack.” Within minutes, the company’s status page confirmed degraded performance across its dashboard, API, and core CDN services.
By 9:30 UTC most services were recovering, and full functionality returned shortly afterward. Canva posted on X at 9:11 UTC that “all systems are now operational,” while Truth Social and other affected platforms came back online without further official statements.
According to Cloudflare’s post-incident report released later Friday, the disruption was caused by a regression introduced during an emergency patch to its Web Application Firewall (WAF) and Bot Management system. Engineers had rushed the update earlier that morning to block active exploitation of a newly disclosed critical vulnerability nicknamed “React2Shell.” Unfortunately, the fix contained a long-dormant bug in the bot-detection feature file that triggered cascading failures across the network when it was deployed globally.
This marks the second significant Cloudflare-related outage in less than three weeks. On November 18, a separate configuration error in the same Bot Management component took major sites offline for more than two hours, affecting Spotify, Discord, ChatGPT, and once again Truth Social.
- The ripple effects on Friday were immediate and far-reaching:
- Graphic designers and marketing teams relying on Canva lost access mid-project, with many taking to X and Reddit to vent frustration.
- Crypto traders reported brief interruptions on centralized exchanges and DeFi front-ends.
- Government websites in Norway and Sweden went dark, temporarily halting public services.
- Corporate users of Zoom, Shopify, and LinkedIn experienced intermittent connectivity issues during peak morning hours in Europe and early business hours in the U.S.

