If users found themselves facing the message “This content is unavailable in this version of the app” while attempting to access ChatGPT on an Android device on the evening of October 9, the issue stemmed from OpenAI. The company promptly advised users to transition to the web version of ChatGPT while they worked to rectify the situation for those on older app versions.
According to OpenAI’s status page, the incident was classified as having a major impact, commencing at 21:00:39Z on October 9 and concluding at 00:26:44Z on October 10, marking a total downtime of three hours and 26 minutes. The information was sourced from the public incident feed, which was reviewed at 21:42 UTC on October 10. OpenAI noted that a fix was being gradually rolled out in the latest Android version, leaving users on outdated builds with limited options.
A second major incident locked admins out of plugin settings
Just 48 minutes after the initial Android issue was logged, OpenAI reported a second major incident. This time, workspace administrators faced challenges in viewing and managing plugin settings, with some users experiencing unavailability of plugins. This incident lasted from 21:48:34Z to 23:26:59Z on October 9, totaling one hour and 38 minutes, with a recovery update issued at 22:52:35Z.
Both incidents overlapped entirely; from 21:48:34Z until the resolution of the admin issue at 23:26:59Z, both problems were active. The Android issue persisted for an additional hour before OpenAI marked it as resolved. Notably, no explanations were provided for either incident, nor was there any indication of how many users were affected.
A third incident was still open on October 10
As of 21:42 UTC on October 10, OpenAI’s status page indicated that while the two major incidents had been resolved, a third, minor incident remained unresolved. This issue, concerning “Delayed Costs data in the Compliance API,” had begun at 17:23:18Z on October 9, well before the Android fault. The latest update regarding this incident was timestamped at 22:03:59Z on October 9.
OpenAI communicated that they had implemented a mitigation strategy and that delayed Costs data was processing through the Compliance API. However, some customers might still experience incomplete Costs data while the backlog was addressed. The company anticipated that this catch-up process would extend through Sunday, October 11. Importantly, ChatGPT usage and API Platform requests remained unaffected, with the overall status designated as “Partial System Degradation” at a minor severity level.
September 29 was longer, and it reached the API
The series of incidents OpenAI faced this autumn has been characterized by variability rather than consistency. A more significant outage occurred on September 29, lasting five hours and 22 minutes and affecting 30 components, far exceeding the duration and scope of the events on October 9. Notably, neither of the two major incidents on October 9 impacted the API, but they did affect two groups that were least equipped to navigate such disruptions: users on outdated Android devices and administrators responsible for managing plugin access.
