Cloudflare’s latest outage briefly knocked parts of the internet offline on Tuesday, blocking access to services including X, ChatGPT, Canva and Grindr for thousands of users worldwide. The San Francisco-based web-infrastructure company, whose network carries roughly a fifth of global web traffic, said it began investigating “internal service degradation” around 6:40 a.m. ET after a spike in unusual traffic at 11:20 UTC triggered widespread 500-style error messages across its network.
Engineers deployed a fix and reported that error levels were falling, though some customers continued to see intermittent problems as systems recovered, and Cloudflare temporarily disabled certain services for users in the U.K. during remediation.
Outage-tracking site Downdetector logged nearly 5,000 problem reports at the peak for Cloudflare and affected platforms, a figure that dropped to about 600 by 8 a.m. ET as access improved.
Cloudflare’s shares slipped about 5% in premarket trading as the company worked to restore full service and emphasized that its teams were “all hands on deck” to ensure traffic flowed normally again.
The incident adds to a string of recent disruptions at major cloud and infrastructure providers, including outages at Amazon Web Services and Microsoft Azure, and underscores how failures at a small number of core internet companies can ripple quickly across social media, gaming, e-commerce and even transit systems.
