DeepSeek, a Chinese startup, announced on Monday that it will temporarily pause new registrations following a cyberattack that coincided with the rapid rise in popularity of its AI assistant.
Earlier in the day, the company’s website experienced outages after its AI assistant achieved the status of the top-rated free application on Apple’s App Store in the United States.
According to the company’s status page, the issues related to its application programming interface and user login problems have been resolved. The outages recorded on Monday were the most significant the company has faced in approximately 90 days, aligning with its newfound popularity.
Last week, DeepSeek introduced a free assistant that promises to use less data at a significantly lower cost compared to existing models. This innovation could signal a shift in the investment requirements for AI technologies.
Utilizing the DeepSeek-V3 model, which its developers claim ranks highly among open-source models and competes with leading closed-source models globally, the AI application has experienced a surge in popularity among U.S. users since its launch on January 10, as reported by app data research firm Sensor Tower.
This achievement underscores DeepSeek’s impact on Silicon Valley, challenging prevailing beliefs about the United States’ dominance in the AI sector and the effectiveness of Washington’s export controls on advanced chips and AI technologies in China.
Technology stocks faced a steep decline on Monday, leading to significant losses for industry giants like Nvidia and Oracle.
AI models, including those from ChatGPT and DeepSeek, rely on advanced chips for their training processes. Since 2021, the Biden administration has broadened bans aimed at preventing the export of these chips to China, where they could be used for training AI models by Chinese companies.
However, DeepSeek researchers indicated in a recent paper that the DeepSeek-V3 utilized Nvidia’s H800 chips for its training, with costs totaling under $6 million.
This claim, although contested, suggests that the chips employed were less advanced than the high-end Nvidia products that U.S. authorities have sought to restrict from export to China. The relatively low training costs associated with DeepSeek have led U.S. tech executives to re-evaluate the effectiveness of the current tech export controls.
Founded in 2023, DeepSeek operates out of Hangzhou and remains relatively unknown. This case marks a notable development since Baidu introduced China’s first large-language model.
While numerous Chinese tech companies have introduced their own AI models since then, DeepSeek is the first to receive acclaim from the U.S. tech community for potentially matching or even surpassing the performance of leading U.S. AI models.
© Thomson Reuters 2025
(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)
