OpenAI’s ChatGPT faces major outage due to power issues in Microsoft data center; now completely restored

ChatGPT experienced a significant outage on Thursday afternoon, leaving many users unable to access the AI ​​chatbot or its associated services. Reports began pouring in around 1:30 PM ET, where users encountered “internal server error” messages. The outage also affected OpenAI’s API and text-to-video generator, Sora, and highlighted these tools’ reliance on seamless backend operations.

OpenAI acknowledged the issue on its status page at 2:00 PM ET and said, “ChatGPT, the API, and Sora are currently experiencing high error rates.” By 6:15 PM ET, the company confirmed that Sora was “fully operational” while APIs began to recover. However, the chatbot itself still faced issues, with engineers continuing to work on a fix.

Although OpenAI did not name the specific “upstream provider” linked to the outage, Microsoft, its exclusive cloud partner, reported a power outage in its South Central US data center. The problem, which started around the same time as ChatGPT’s downtime, caused storage lag, timeouts, and HTTP 500 errors. Microsoft’s cloud gaming service, Xbox Cloud, was also affected.

By 5:00 PM ET, Microsoft reported that it had “fully restored” power to the affected data center, but the incident underscores the cascading effects that infrastructure failures have on dependent services.

This is not the first time ChatGPT has gone offline in recent months. Earlier in December, ChatGPT and Sora faced extended downtime shortly after the video generator was launched to subscribers. In June, a widespread outage disrupted several AI tools, including ChatGPT, highlighting the challenges of scaling services that rely on complex cloud infrastructure.