ChatGPT users were reporting widespread problems Thursday, with outage reports on Downdetector surging past 35,000 in what appeared to be another disruption affecting OpenAI’s chatbot.
Users reported difficulty accessing or using ChatGPT, only days after another service incident affected the platform.
ChatGPT users report widespread problems days after major OpenAI outage
(Photo: Meir Chaimowitz/shutterstock)
OpenAI’s official status page, however, was not reporting an active outage at the time of publication and said its systems were “fully operational.”
The discrepancy is not unusual during fast-moving service disruptions: user-reporting platforms such as Downdetector can register spikes before a company confirms an incident, and reports do not necessarily mean every user is affected.
The apparent disruption comes after OpenAI acknowledged high error rates in ChatGPT Work Mode earlier Thursday. That incident began at 12:04 a.m. UTC and was declared resolved six minutes later, with OpenAI saying all affected services had fully recovered.
OpenAI has recorded several other service problems in recent days. On Wednesday, it resolved elevated errors affecting new account creation, while on Monday ChatGPT Work experienced elevated errors and latency. The company also reported elevated conversation errors affecting Free and Go users earlier this week.
During Monday’s incident, some users were unable to start or continue tasks while OpenAI worked on a mitigation.
OpenAI notes that its availability figures are aggregated across subscription tiers, models and error types, meaning individual users can experience problems even when overall service availability appears normal.
The cause of Thursday’s latest spike in user reports was not immediately clear, and OpenAI had not announced a new incident on its status page at the time of publication.