On November 18, 2025 (local time), Cloudflare, a global behemoth in internet infrastructure, suffered a colossal service outage. This left numerous major websites around the world completely inaccessible. Data from Downdetector, a website outage - tracking platform that was also briefly affected, revealed that platforms like Anthropic's Claude chatbot, Trump's Truth Social, Musk's social media platform X, and some digital services of New Jersey's public transportation system were impacted. OpenAI's status page showed that ChatGPT and its Sora short - video app experienced outages due to issues with a "third - party service provider," but they have since made a full recovery.

Cloudflare was founded at Harvard University in 2009. It rolled out its first beta version in 2010 and went public on the New York Stock Exchange in 2019. At present, it caters to 30% of Fortune 1000 companies, and its traffic management and security protection services handle roughly 20% of global internet traffic.

This incident took a toll on Cloudflare's stock price, which dropped by 2.83% by the close of U.S. markets on the 18th. Matthew Prince, the co - founder and CEO of the company, admitted that this was Cloudflare's most severe outage since 2019. He apologized on behalf of the team for the disruptions it caused to the internet.

Dane Knecht, the CTO, posted on social media that the incident was triggered by a latent defect in a service supporting the company's bot mitigation function. After a routine configuration change, this defect caused the service to crash, leading to a widespread degradation of network and other services. It was not the result of an attack.

Early in the morning of November 19 (local time), Cloudflare released a comprehensive report detailing the nearly 5 - hour event. At 11:28 a.m. on the 18th, the impacts started to surface, with errors first detected in customer HTTP traffic. By 2:30 p.m., the primary impacts had been resolved. Errors in downstream - affected services decreased, and most services began to function correctly. By 5:06 p.m., all downstream services had been restarted, and operations had fully resumed.

Cloudflare stated that it has already started looking into ways to fortify its systems to prevent similar failures in the future. This includes measures such as improving the ingestion processing of configuration files and enabling more global emergency stop switches for functions.