On Thursday afternoon, ChatGPT, the popular AI chatbot developed by OpenAI, experienced an unexpected and significant interruption of service. Worldwide, users encountered difficulties accessing it, and many received “internal server error” messages.
The AI chatbot stopped responding to queries, marking the start of a significant outage. At that time, people were unable to access the chatbot.
The outage of ChatGPT started around 1:30 PM ET. After that, at 2:00 PM ET, OpenAI acknowledged the issue with ChatGPT, its API, and Sora (ChatGPT’s platform for creating videos from text prompts).
They stated that they were “currently experiencing high error rates.”
Investigations into this issue revealed that the root cause of the disruption was a power outage in Microsoft’s data center in the South-Central US region. This data center is OpenAI’s exclusive cloud provider. Although OpenAI didn’t specifically mention the root cause, Microsoft did.
Microsoft reported that the outage, which began at 18:44 UTC (1:44 PM ET), caused storage delays, timeouts, and HTTP 500 errors when the users tried to access this service.
By 5:00 PM ET, Microsoft announced the restoration of power to the impacted data center and continued to validate recovery efforts. While the Sora and APIs of OpenAI began recovering by 6:15 ET, ChatGPT’s restoration took a little longer.
This is not the first time ChatGPT has experienced significant downtime. It happened earlier this month and as well as in July.
The latest outage underscores AI services’ critical dependence on infrastructure and highlights the impact of unexpected events. Despite its occasional downtime, ChatGPT remains a widely used tool.
