ChatGPT users faced major disruptions Monday as at least 12 sub-services showed degraded status simultaneously. The incident left many unable to load essential features like login, search, image generation, and voice mode. OpenAI confirmed the system failure but has not disclosed the total user impact.
The primary disruption began at 12:43 PM when OpenAI officially logged the incident. At the time of writing, the outage had lasted roughly two hours. The company’s status page said: “We are continuing to investigate the issue for the listed services.”
While the outage prevented users from loading the API Platform web portal, OpenAI’s API endpoints showed no disruption on the status page. Third-party applications use these endpoints to integrate OpenAI’s models into their own products.
The disruption also affected the Codex app, showing a degraded status for users. However, four other Codex services continued to operate without interruption. These active developer tools included the web interface, Codex API, CLI, and VS Code extension.
Business accounts face separate errors
A separate incident, which began roughly nine hours earlier, impacted ChatGPT Business accounts attempting to upgrade plans or add seats. Users faced access failures lasting up to an hour. The status page for this incident listed no components as officially affected, despite the access failures reported by users.
OpenAI applied a technical mitigation for these business accounts at 3:13 PM. The company has not indicated whether the two incidents share a common cause. OpenAI stated: “We have applied the mitigation and are monitoring the recovery.”