Cloudflare outage takes X, ChatGPT and other major sites offline

A major Cloudflare outage has triggered a global internet disruption, knocking offline some of the world’s most heavily used websites.

The likes of X, ChatGPT, Canva and major online services have gone dark, in what has quickly become one of the largest internet failures of the year.

The issue began on Tuesday morning and escalated rapidly, with Cloudflare confirming a “global network problem” affecting vast parts of its infrastructure.

Users across the UK, US, Europe and Asia reported being unable to load timelines on X, run prompts in ChatGPT or access everyday tools like design platforms, betting sites and business services.

Downdetector recorded thousands of outage reports within minutes, with X peaking at nearly 10,000 alerts as users were repeatedly met with Cloudflare error pages instead of platforms loading normally.

Cloudflare, one of the internet’s most critical backbone providers, said engineers are “continuing to investigate the issue” and stressed there is no evidence of a security breach, adding that user data remains safe.

The company is also carrying out planned maintenance across several data centres, though it has not yet confirmed whether the work contributed to today’s mass outage.

One of the world’s largest edge networks

As one of the world’s largest edge networks, Cloudflare handles traffic for millions of websites, keeping them fast, secure and shielded against attacks.

But today’s malfunction shows how deeply intertwined the modern internet has become.

The scale mirrors a previous Amazon Web Services (AWS) outage in October that brought down hundreds of platforms simultaneously, but today’s impact appears even broader.

Organisations have begun issuing internal advisories, warning staff of delays and degraded performance across digital systems until Cloudflare restores stability.

Some websites have begun to flicker back online, though users continue to report instability and slow recovery.

Cloudflare has not provided a timeline for a full fix, and the outage is still unfolding.

Related posts

United Against Online Abuse Welcomes 5th Scholar to Fully Funded Research Programme

No selfies please: Croatia has a quiet luxury island that’s more Succession than Kardashian

Fitch Learning Completes Acquisition of Moody’s Analytics Learning Solutions and the Canadian Securities Institute