Getty Images/iStockphoto
Risk & Repeat: DeepSeek security issues emerge
The introduction of DeepSeek's new generative AI models have been met with fervor, but security issues have created apparent challenges for the Chinese startup.
DeepSeek's exploding popularity has been met with growing interest -- and concerns -- from security researchers.
The Chinese AI startup gained a massive attention over the last week following the Jan. 20 introduction of its first its first-generation large language models, DeepSeek-R1-Zero and DeepSeek-R1. DeepSeek drew quick interest for the capabilities of its AI reasoning models as well as the reportedly low cost to train it, such that it overtook ChatGPT on the Apple App Store's "Free Apps" chart and had major impact on AI-relevant tech stocks like Nvidia.
But in addition to the overwhelming discourse surrounding DeepSeek's introduction, the startup has also seen its share of security issues. On Monday DeepSeek posted to its status page that due to what it called "large-scale malicious attacks" against its services, it temporarily limited new registrations to its platform. DeepSeek has not clarified the nature of the attacks, and registration for new accounts has apparently resumed.
Additionally, cloud security vendor Wiz published research Wednesday regarding an exposed DeepSeek database containing "over a million lines of log streams containing chat history, secret keys, backend details, and other highly sensitive information." Wiz said it responsibly reported the exposure to DeepSeek, which promptly addressed the issue. But other security researchers reported similar vulnerabilities as well as risks like jailbreaking.
Informa TechTarget editors Rob Wright and Alex Culafi discuss the latest news surrounding DeepSeek's security issues on the Risk & Repeat podcast.
Alexander Culafi is a senior information security news writer and podcast host for TechTarget Editorial.
