Recently, OpenAI's AI chatbot ChatGPT experienced service outages, causing some users to be unable to use it normally on December 2nd and 3rd. According to user feedback, this failure mainly affected the web version of ChatGPT, with many users finding that the service was unresponsive or failed to load when trying to interact through their browsers.

Unlike the issues with the web version, the Mac desktop client performed normally during this period, responding smoothly to user commands. This phenomenon sparked speculation about the cause of the failure, which seemed to stem from OpenAI's web servers or related network infrastructure. 

image.png

According to data from OpenAI's official status page, there were indeed periods of instability on December 2nd and 3rd. The first failure occurred at 2:08 AM Beijing Time on December 2nd, and OpenAI took mitigation measures at 2:37 AM and monitored the recovery. By 3:00 AM that day, the service had returned to normal. However, the problem was not over, as another failure occurred on December 3rd, leading to an "increase in ChatGPT error rates," which started at 3:26 AM and was fully resolved by 4:12 AM.

In a statement to CNBC, OpenAI explained: "Some users encountered issues while using ChatGPT due to a routing configuration error. This issue has now been fixed." According to statistics from the service outage tracking website Downdetector, approximately 3,000 users reported problems with the chatbot on Tuesday. Although this outage caused inconvenience for some users, the impact was relatively limited and did not spread to all users or platforms.

After emergency repairs, ChatGPT has now restored normal service. Users can once again use this powerful AI assistant to enjoy a smoother chatting experience.