OpenAI Discloses Rogue Incidents
OpenAI has disclosed four new incidents, affecting their services. No specific technical details or affected versions are provided in the reports.
since
Here's why. Incidents from Saturday, October 17, 2026 (UTC), ranked byBreakage Velocity.What is Breakage Velocity?A live score of how fast an outage or regression is spreading. The more people hit it, and the faster, the higher it climbs; it cools off as things calm down.Open-ended scale: under 25 low · 25+ elevated · 100+ high · 300+ extreme.
8 outages, 2 still active today. Worst: Fly.io (9 h 9 min).
OpenAI has disclosed four new incidents, affecting their services. No specific technical details or affected versions are provided in the reports.
since
A platform-wide rolling host reboot incident affecting all Fly.io users. No official fix URL provided.
– UTC
Fireworks AI experienced a service degradation affecting the Qwen 3.8 Max model. The incident has been resolved and the service is fully operational.
– UTC
A third-party service used by Netlify's build system caused increased errors for builds using the CSP extension. The issue has been resolved.
– UTC
Fireworks AI experienced a service degradation affecting Qwen 3.8 Max. The incident has been resolved and the service is now operational.
– UTC
OpenRouter experienced delays affecting activity logs and usage metrics reporting. Core inference functionality remained unaffected throughout the incident, and the issue has since been resolved.
– UTC
Project creation and unpausing errors occurred in the ap-northeast-1 region. The issue has been resolved and services are fully restored.
– UTC
The kubelet crashes and panics with an interface conversion error when a PodCertificateRequest deletion arrives as a tombstone after a relist. Users running affected versions experience kubelet instability.
since