DeepSeek, the rising star in China's artificial intelligence (AI) sector, has temporarily restricted new user registrations following a wave of malicious cyberattacks targeting its platform.
"Due to large-scale malicious attacks on DeepSeek's services, we are temporarily limiting registrations to ensure continued service," the company announced on its incident report page. "Existing users can log in as usual. Thank you for your understanding and support."
When attempting to register, new users are met with a notification that states, “Registration may be busy,” encouraging them to wait and try again later.
Growing Popularity Attracts Cyber Threats
Eric Kron, a security awareness advocate at KnowBe4, commented on the incident:
“With DeepSeek's growing popularity, it’s no surprise that the platform is becoming a target for malicious web traffic. These attacks could be attempts to extort the company, sabotage by competitors, or even actions by investors seeking to protect rival organizations.”
Founded in 2023, DeepSeek has gained international recognition for its commitment to making artificial general intelligence (AGI) a reality. The company recently achieved a major milestone when its iOS chatbot app surpassed OpenAI’s ChatGPT to claim the #1 spot on Apple’s Top Free Apps chart in the U.S.
Advanced AI Models at a Fraction of the Cost
DeepSeek has released a series of innovative reasoning and mix-of-experts language models under an MIT license. The company claims these models outperform Silicon Valley competitors while being significantly cheaper to train, even amidst U.S. sanctions restricting the sale of advanced AI chips to Chinese firms.
According to DeepSeek, its flagship model, DeepSeek-V3, achieves remarkable efficiency during training:
- Training 1 trillion tokens requires only 180K H800 GPU hours (equivalent to 3.7 days on a 2,048 H800 GPU cluster).
- The entire pre-training process is completed in under two months, costing just 2.788M GPU hours or $5.576M at $2 per GPU hour.
Controversies Surrounding Content Censorship and Data Privacy
Despite its technological breakthroughs, DeepSeek has sparked criticism over its censorship policies. The platform restricts responses to politically sensitive topics, including Tiananmen Square, Taiwan, and the treatment of Uyghurs in China.
Additionally, DeepSeek's privacy policy raises concerns by hosting user data—such as device details, network information, and payment records—on servers located in the People’s Republic of China. These practices could further strain U.S.-China relations, particularly amid ongoing scrutiny of Chinese-owned apps like TikTok.
A New Era of Global AI Competition
The rise of DeepSeek has introduced fresh competition into the global AI landscape, drawing praise from industry leaders. NVIDIA’s senior research manager, Jim Fan, remarked, "We are living in a timeline where a non-U.S. company is keeping the original mission of OpenAI alive—truly open, frontier research that empowers all."
Even OpenAI CEO Sam Altman acknowledged the quality of DeepSeek's R1 reasoning model, calling it “impressive” and expressing excitement over the emergence of a new competitor.
Conclusion
As DeepSeek grapples with the challenges of scaling its services amid cyberattacks, the company’s rapid rise and innovative approach to AGI development solidify its position as a key player in the AI ecosystem. However, ongoing concerns over censorship and data privacy may shape its future trajectory on the global stage.
