Overview
- OpenAI's status page acknowledged increased errors starting around 11 a.m. ET and said it was implementing a mitigation.
- By about 12:35 p.m. ET, the status page listed ChatGPT as fully operational, indicating the incident had cleared.
- DownDetector showed a sharp surge in user reports, with close to 3,000 in the U.K., and social posts described slowdowns or failures to load responses.
- Monitoring pages and users also noted concurrent spikes for other services, including AWS, Grok, and NordVPN, with Grok later reporting normal operations.
- Reporters and users floated possibilities such as a cloud-provider issue or the recent Atlas rollout, but no confirmed root cause has been provided.