DeepSeek’s chatbot was down for more than seven hours overnight in China, forcing the AI pioneer to deploy several updates to rectify the issue. The chatbot users started complaining about faults on Sunday evening, according to Downdetector. The startup’s own status page acknowledged an initial issue at 9:35 p.m. before marking the incident resolved two hours later.
Following updates on Monday showed DeepSeek addressing another case of performance issues that took until 10:33 a.m. to be fixed.
ALSO READ: West Asia War Impact: Rupee Breaches 95-level Against US Dollar For First Time; Key factors
The causes of the outages is yet to be ascertained and DeepSeek did not immediately respond to an emailed request for comment.
The extended downtime is unexpected for a globally used app like DeepSeek, which has for over a year carried the mantle of China’s breakthrough artificial intelligence service. It’s also uncommon for DeepSeek itself, which has maintained close to a 99 per cent operational record since it unveiled the popular R1 in January 2025, according to its status page.
Speculation has spread since the start of the year that Hangzhou-based DeepSeek is preparing to roll out a major update, after its initial splashy debut on Jan. 20 last year. That prompted local rivals from Alibaba Group Holding Ltd. to ByteDance Ltd. and Tencent Holdings Ltd. to release a barrage of new AI models and services over the Lunar New Year holiday. Anticipation for DeepSeek’s next big move remains high, though the publicity-shy company has kept mum on a timeline.
Several Controversies Regarding DeepSeek
DeepSeek's terms of service explicitly say that it will gather and use user interactions, in contrast to certain AI companies that provide options for user privacy. This raises questions about who can access this data, how it is preserved, and whether it might be shared with other parties, including governmental organisations.
Given that DeepSeek was created in China and the ongoing rivalry between China and the US, it is reasonable enough to be concerned that the model might be used for spying and mass data collecting. Furthermore, DeepSeek might be used as a tool for state-sponsored propaganda, because AI models have the power to influence public opinion.
Responsible data sourcing is necessary for ethical AI research, but DeepSeek is unclear about the legality of its training data acquisition and its adherence to intellectual property rights.