DeepSeek is the target of a cyberattack.

(Reuters) Chinese startup DeepSeek announced on Monday that it will temporarily limit new sign-ups due to a cyberattack, after the company's artificial intelligence assistant gained sudden popularity.

Earlier in the day, the startup also faced outages on its website after its AI assistant became the highest-rated free app on Apple's App Store in the United States.

According to its status page, the company resolved issues related to the application programming interface and users' inability to log in to the website. Monday's outages were the company's longest in about 90 days, coinciding with an exponential increase in its popularity.

Last week, DeepSeek launched a free assistant that, according to the company, uses less data and operates at a fraction of the cost of competing AI models, possibly marking a turning point in the investments needed for AI development.

Powered by the DeepSeek-V3 model, which, according to its creators, "leads the rankings among open-source models and rivals the world's most advanced closed-source models," the artificial intelligence application has gained enormous popularity among users in the US since its launch on January 10, according to data from the research firm Sensor Tower.

This milestone highlights how DeepSeek has had a profound impact on Silicon Valley, challenging the widely accepted view of US primacy in AI and the effectiveness of Washington's technology export controls targeting China's advanced chip and AI capabilities.

Shares of technology companies suffered a sharp decline on Monday, with Nvidia and Oracle stocks registering significant losses.

AI models like ChatGPT and DeepSeek require advanced chips to power their training. Since 2021, the Biden administration has broadened the scope of bans to prevent these chips from being exported to China and used to train AI models by Chinese companies.

However, researchers at DeepSeek wrote in a paper last month that DeepSeek-V3 was trained using Nvidia's H800 chips, which cost less than $6 million.

Although this claim has been disputed, the fact that the chips used are less advanced than the more sophisticated models that Washington is trying to prevent from reaching China, as well as the relatively low training costs, has led US technology executives to question the effectiveness of technology export controls.

Little is known about the company behind DeepSeek, a small startup based in Hangzhou and founded in 2023, the same year that Chinese search giant Baidu launched China's first large-scale language model.

Since then, dozens of Chinese technology companies, large and small, have launched their own AI models, but DeepSeek is the first to be praised by the US technology industry as being able to match or even surpass the performance of the most advanced models developed in the United States.