Cloudflare Outage Disrupts X, ChatGPT, and Major Sites

A widespread Cloudflare outage briefly broke large parts of the internet today, sending users into a maze of error pages across some of the world’s biggest apps and sites. X (formerly Twitter), ChatGPT, Letterboxd, bet365 and even outage tracker DownDetector were among services showing “internal server error” style messages tied to Cloudflare’s network issues.

Cloudflare’s status page described the incident as an “internal service degradation” affecting multiple products across its global network, including Access, WARP, CDN/Cache, Firewall, Workers and the dashboard. The company said it identified the problem early afternoon UTC and began rolling out a fix. By 14:42 UTC it reported a remedy had been implemented and moved into monitoring, while cautioning that some users could still see residual issues, particularly with dashboard logins. Temporary steps included disabling WARP access in London before re-enabling it as recovery progressed.

From a user’s perspective, the symptoms were confusing first security-style blocks, failed page renders, and intermittent loading on sites that normally sit behind Cloudflare’s performance and protection layers. That’s the nature of a backbone issue at a company of Cloudflare’s scale. When the network that accelerates and shields a big slice of the web stumbles, the ripple hits everything from social feeds to streaming, shopping checkouts and developer tools in one go. Reports of broken functionality began stacking up late morning to midday UTC, and The Independent’s live blog documented popular sites blinking in and out until Cloudflare’s fix began to stick.

Cloudflare hasn’t posted a full post-mortem yet, but the company’s incident log makes clear the problem originated inside its own services, not with any single customer or a broad external attack. The sequence of updates shows the team moving from investigation, to identification, to mitigation, then to an all-clear with continued monitoring. That aligns with what users saw: a rough first hour, partial recoveries, and then a general return to normal.

As of this writing, Cloudflare says the issue is resolved and services are back to expected levels, with engineers keeping watch for stragglers. If your favourite app briefly told you that you were a “security risk” or simply refused to load, it wasn’t you, your device, or your ISP, it was an upstream problem in the fabric a lot of the internet runs on, and it’s been fixed.