OpenAI confirmed a partial outage on August 31, 2026, affecting ChatGPT users worldwide with network errors and latency issues.
OpenAI confirmed a partial outage on August 31, 2026, that caused network errors and slower response times for a subset of ChatGPT users worldwide.
Outage timeline and scope
The issue began early on August 31, when users reported receiving "network error" messages while attempting to submit prompts. The problem persisted for several hours, with latency spikes affecting both the free and paid tiers of the service.
OpenAI’s status page indicated that the disruption was limited to specific data center clusters, rather than a complete platform failure. Engineers worked to isolate the faulty network routes and restore normal performance.
Technical cause
According to the company’s brief technical note, a misconfiguration in the load‑balancing layer caused packets to be dropped intermittently. This resulted in the error messages users saw and the noticeable lag in generating replies.
The misconfiguration was unrelated to any external cyber‑attack, and no user data was compromised during the incident.
Impact on users
Affected users experienced delayed answers, incomplete responses, or outright failure to receive a reply. The outage was reported across multiple regions, including North America, Europe, and parts of Asia.
- Increased response times up to 30 seconds
- Frequent "network error" prompts
- Temporary unavailability of new chat sessions
OpenAI restored full functionality by early afternoon UTC, and the status page was updated to reflect normal operations. The company apologized for the inconvenience and promised a post‑mortem analysis.
Comments
No comments yet.