Chinese artificial intelligence startup DeepSeek announced on Monday that it will temporarily limit new user registrations following a cyberattack, just as its AI assistant experiences a meteoric rise in popularity.

Earlier in the day, DeepSeek’s website suffered outages after its AI assistant became the top-rated free app on Apple’s U.S. App Store, leading to a surge in traffic. According to the company’s status page, issues related to its API and user logins were quickly addressed. However, Monday’s disruption marked the longest downtime DeepSeek has experienced in the past three months, coinciding with its rapid ascent in the AI space.

The startup recently introduced a free AI assistant that it claims operates with significantly lower data usage and at a fraction of the cost of competing models. This development could signal a new shift in the AI industry, demonstrating that high-performance AI can be achieved with lower investment.

A Breakthrough in AI Efficiency

DeepSeek’s rise has been fueled by its DeepSeek-V3 model, which the company claims outperforms all open-source AI models and even rivals leading closed-source models worldwide. Since its release on January 10, the AI assistant has quickly gained traction among U.S. users, according to data from Sensor Tower.

This rapid success has challenged traditional assumptions about U.S. dominance in AI and raised new concerns about the effectiveness of Washington’s export controls aimed at restricting China’s access to advanced AI and semiconductor technologies.

The news also had an immediate impact on the stock market, with tech giants like Nvidia and Oracle seeing their shares take a hit on Monday.

U.S. Tech Sanctions Under Scrutiny

Like other cutting-edge AI models, DeepSeek relies on powerful chips for training. Since 2021, the Biden administration has expanded restrictions on advanced semiconductor exports to China, aiming to prevent their use in training Chinese AI systems.

However, DeepSeek researchers revealed last month that they trained DeepSeek-V3 using Nvidia’s H800 chips, reportedly spending under $6 million—a significantly lower cost than what is typically required for developing high-performance AI models.

While some experts dispute this claim, the revelation has sparked debate among U.S. tech leaders about whether current export controls are truly effective in limiting China’s AI advancements.

A Rising AI Contender from China

DeepSeek, a relatively unknown startup based in Hangzhou, was founded in 2023, the same year search engine giant Baidu launched China’s first large-scale AI language model. Since then, numerous Chinese companies have entered the AI race, but DeepSeek stands out as the first to receive significant recognition from the U.S. tech industry, with some experts suggesting its model matches—or even surpasses—the performance of leading American AI systems.

As DeepSeek continues to gain momentum, its success could reshape the global AI landscape and further intensify competition between Chinese and U.S. AI firms.

Source