The failure cascaded through essential online services, serving as a reminder of the fragility of interconnected internet ecosystems and dependency risks.
A widespread internet outage caused by a major internet infrastructure provider, is disrupting access to numerous popular platforms as of 18 Nov 2025 (10pm Singapore time).
Adding to four earlier outages this year, the current outage by Cloudflare is affecting X (formerly Twitter), ChatGPT, Canva, and several cryptocurrency services. The incident has already resulted in thousands of users experiencing intermittent connectivity problems, with many encountering “500 internal server errors” that have rendered significant portions of the web inaccessible for hours.
“Internal service degradation” cited
The outage began early Tuesday (UTC) when Cloudflare’s systems were deemed to have suffered what the firm called “internal service degradation.” This affected its core dashboard, its API, and triggered widespread errors for its clients, many of whom power mainstream applications and services worldwide.
The firm has acknowledged the issue on its status page, and indicated ongoing investigations, though initial recovery efforts periodically stalled as problems persisted. Other platforms such as Spotify, Grindr, and several gaming and crypto-related sites were inaccessible or unstable during the outage.
The disruption had extended to major crypto exchanges and services confirming downtime linked to Cloudflare’s troubles.
Notably, the crypto sector’s reliance on Cloudflare was underscored as both trading and blockchain explorer sites showed persistent errors.
Technical outage details
Cloudflare’s engineers initially pointed to scheduled maintenance at its Santiago, Chile data center, but the problem quickly expanded beyond a single location, suggesting a more systemic failure.
The recurring “500” errors signaled general internal server problems, meaning affected websites could not process requests as Cloudflare’s routing and protection services faltered.
The outage is demonstrating the internet’s dependence on centralized infrastructure providers: one single point of failure can cascade into multiple failures rippling through thousands of independent websites and applications globally. Only recently, the world was served a reminder by a significant Amazon Web Services outage, raising growing concerns about the single points of failure in cloud-based and web service ecosystems.
As service gradually resumes, the firm is continuing investigations into the root causes of this fifth consecutive outage in 2025:
- 21 August, 2025: Network congestion incident triggered by traffic spikes in AWS us-east-1 region causing disruptions on the firm’s network links, impacting service reliability
- 14 July, 2025: Outage of the firm’s 1.1.1.1 DNS resolver lasting about 62 minutes caused by an internal configuration error, affecting DNS resolution for users worldwide
- 12 June, 2025: Major outage impacting multiple Cloudflare services including Workers KV, WARP, Access, Gateway, Images, Stream, with root cause traced to third-party cloud storage infrastructure failure, lasting over 2 hours
- 21 March, 2025: Global outage from credential rotation errors leading to widespread failures in the firm’s service write and read operations
