Cloudflare outage caused by botched blocking of phishing URL

Cloudflare experienced a 59-minute outage across R2 Object Storage and dependent services after an employee, responding to an abuse report, mistakenly disabled the R2 Gateway instead of a specific endpoint. The incident caused complete failures in Stream, Images, Cache Reserve, Vectorize (for certain operations), and other services, along with partial disruptions to additional systems and some log delivery data loss. Cloudflare attributed the event to human error and insufficient safeguards, implemented immediate restrictions on high-impact actions, and plans enhanced access controls and two-party approvals; the report also references a prior November 2024 incident that caused significant log loss.