China's artificial intelligence chatbot DeepSeek has just experienced a shutdown lasting more than 7 hours, considered the longest since the platform became popular in early 2025.
According to the company's official status page, the system encountered a "serious shutdown" error from dawn on Monday and was only fixed at about 10:33 am on the same day (local time), equivalent to 7:13 am continuous interruption.
This is a rare time that a service for general users has been affected for such a long time.
DeepSeek did not give any explanation for the cause of the incident. However, experts believe that such disruptions may stem from many factors, including server errors, system overload, or problems during software updates.
Previously, at the end of January 2025, DeepSeek's API service, which is a tool for developers to integrate chatbots into applications, also encountered a problem lasting nearly a day, right at the time when this platform attracted a sharp increase in users.
However, for users accessing the web directly, previous incidents rarely lasted longer than 2 hours.
This incident takes place in the context that the global technology industry is following DeepSeek's next step. After the success of the R1 and V3 models, the company is expected to soon launch a new generation of AI with superior capabilities. However, so far, DeepSeek has not announced a specific time.
The prolonged disruption shows the infrastructure challenges that AI platforms are facing as user size increases rapidly.
At the same time, it also sets higher requirements for information stability and transparency for AI services that are playing an increasingly important role in digital life.
