China’s AI start-up DeepSeek, a prominent platform specializing in advanced data analytics, has temporarily suspended new user registrations after “large-scale malicious attacks” targeted its systems.
Founded in 2023, DeepSeek is a Chinese company “dedicated to making AGI [artificial general intelligence] a reality.” The AI start-up gained worldwide attention this week when its AI app for iOS, DeepSeek—AI Assistant—became the highest-rated free app on Apple’s App Store in the United States, briefly surpassing OpenAI’s ChatGPT in popularity.
Talking about the cybersecurity incident late Monday, the company termed it as a “deliberate and sophisticated breach” aimed at compromising its infrastructure.
While DeepSeek highlighted that no customer data had been accessed or compromised, the cyber-attack disrupted its registration systems, prompting a precautionary halt to onboarding new users.
“Due to large-scale malicious attacks on DeepSeek’s services, we are temporarily limiting registrations to ensure continued service. Existing users can log in as usual. Thanks for your understanding and support,” the company said on the status page.
When attempting to register for DeepSeek’s services, users encounter an error message on the sign-up page stating that new registrations are temporarily unavailable due to malicious attacks. The message advises them to wait and try again. Some users also see a similar message indicating that “registration may be busy.”
DeepSeek’s security team is assessing the extent of the cyberattack. It is also collaborating with external experts to investigate the incident and strengthen protective measures. As the investigation progresses, the team plans to provide further updates in the coming days.
“We are prioritizing security enhancements and will restore full functionality once we are confident in the robustness of our defenses,” a company spokesperson said.
Existing users remain unaffected and can continue logging in without disruption. The platform’s core services are operating as usual. However, the company has advised its users to stay vigilant, update their passwords, and enable multi-factor authentication.
For now, new users will need to wait until the platform ensures its security measures can withstand similar threats in the future.
Last week, the Chinese start-up released its new free, open-source reasoning model, DeepSeek-R1, designed to rival OpenAI’s o1, which went on to quickly gain global attention for its significant low cost and computing power.
The groundbreaking DeepSeek-V3 model, with over 600B parameters, powers the AI assistant, which the company calls “a leading open-source model that can compete with the world’s most advanced closed-source models.” According to the developers, the DeepSeek-V3 model was trained on less powerful Nvidia H800 chips, costing under $6 million.
