A widespread internet outage on Tuesday morning temporarily took down or impaired major services including ChatGPT, Claude, Spotify, and X, after a critical failure at Cloudflare, one of the world’s largest internet infrastructure providers. By 8 a.m. ET, Cloudflare said it had identified the issue and began deploying a fix, later confirming that the incident had been resolved within two hours, though monitoring would continue to ensure full stability.
Cloudflare’s chief technology officer, Dane Knecht, attributed the disruption to a latent bug in a system supporting the company’s bot mitigation tools. He explained that the bug surfaced following a routine configuration change and triggered cascading failures across Cloudflare’s global network. Emphasising that the outage was not the result of an attack, Knecht apologised to customers and the broader internet community, acknowledging the disruption caused and vowing to prevent such failures in the future.
Despite the rapid recovery, Cloudflare noted that some users continued to encounter login and dashboard issues. The firm said it was working to apply a separate fix to fully restore functionality and would maintain heightened monitoring for any additional problems. Knecht also promised a detailed technical breakdown of the incident to be published later.
The outage highlights the fragility of the modern web and its dependence on a small cluster of infrastructure providers. Cloudflare alone supports an estimated 20% of all websites, operating data centres in 330 cities and connecting with 13,000 networks globally, including major ISPs and cloud providers. The incident, coming less than a month after a similar disruption at AWS, underscores how failures at any of these backbone companies can ripple across the internet an irony made sharper by Cloudflare’s core mission to protect sites from outages and DDoS attack