Cloudflare outage knocks major platforms offline as company blames latent bug and issues apology

A major internet disruption unfolded on Tuesday morning after a widespread Cloudflare outage caused several popular services to go down or respond sluggishly. Users experienced issues across ChatGPT, Claude, Spotify, X and dozens of other platforms, all of which rely on Cloudflare’s infrastructure to stay online.

Cloudflare acknowledged the problem on its status page shortly after 8 a.m. ET, confirming that the incident had been identified and a fix was being deployed. Within two hours, the company announced that the issue had been resolved and services were recovering, although monitoring would continue.

In a post on X, Cloudflare chief technology officer Dane Knecht explained that the root cause was a “latent bug” that had silently existed within one of the firm’s internal systems. This type of bug remains dormant in normal conditions and escapes detection during routine testing, but can suddenly trigger a failure when a specific change is made.

Knecht revealed that this hidden flaw surfaced after a routine configuration update, which caused part of Cloudflare’s bot mitigation system to crash. That single failure set off a chain reaction that rippled through the company’s network and affected multiple services worldwide.

“In short, a latent bug in a service underpinning our bot mitigation capability started to crash after a routine configuration change we made. That cascaded into a broad degradation to our network and other services. This was not an attack,” he wrote.

Knecht apologised to customers, stating that Cloudflare had let them down along with “the broader internet.” He added that the company is working to ensure such an incident does not happen again and promised a detailed postmortem soon. “I know it caused real pain today,” he said.

Even as services largely returned to normal, Cloudflare noted that some users might still face issues when accessing the Cloudflare dashboard. Engineers are working on a separate fix for this while continuing to track any lingering abnormalities.

The incident arrives only weeks after an outage at Amazon Web Services, underscoring how dependent the modern internet has become on a small group of infrastructure providers. Estimates suggest Cloudflare supports around 20 percent of all websites, with data centres in 330 cities and direct connections to 13,000 networks worldwide. The company is also a major provider of protection against Distributed Denial of Service attacks, making Tuesday’s disruption particularly striking.