Cloudflare says its systems are gradually recovering after a major global outage on Tuesday that knocked several top websites offline for hours.
Many users, however, are still experiencing slow responses, error messages and unstable access as the company works on full restoration.
The disruption began around 11:48 UTC on November 18, 2025. It affected platforms like X, ChatGPT, Spotify, Canva, Letterboxd, Shopify and Indeed. Across Nigeria, many major websites were also unreachable for hours.
OpenAI reported that both ChatGPT and its Sora video tool were experiencing problems linked to “an issue with one of our third-party service providers.”
A Cloudflare spokesperson said the company detected a sudden rush of strange traffic around 6:20 a.m. ET that led to errors across its network. The spokesperson said, “We do not yet know the cause of the spike in unusual traffic. We are all hands on deck to make sure all traffic is served without errors.”
In an update posted at 9:22 a.m. ET, the company said it was “continuing to work on a fix.”
Cloudflare’s technology supports about 20 per cent of global web traffic. It helps protect sites from cyber attacks and keeps them stable during heavy demand. Because of its wide coverage, outages often have far-reaching impact.
Downdetector, the site that tracks outages, also had access problems. NJ Transit reported delays in some digital services “due to a vendor.”
Other affected platforms include Anthropic’s Claude chatbot and Truth Social.
Cloudflare’s stock dropped more than 3 per cent following the outage.
This disruption comes shortly after Amazon Web Services suffered a long service breakdown, and weeks after Microsoft’s Azure and 365 services faced a global outage.
Last year, a faulty software update from cybersecurity company CrowdStrike also triggered widespread chaos, grounding flights and affecting hospitals and financial institutions.
