Deepseek AI Exposes Sensitive Data Online, Says Wiz
New York-based cybersecurity firm Wiz has reported that Chinese artificial intelligence (AI) company Deepseek inadvertently left a vast amount of sensitive data exposed online. The revelation was reported by Reuters.
In a blog post published on Wednesday, Wiz stated that their scans detected over a million lines of unprotected data from Deepseek. The exposed information included software keys and user chat logs, which stored prompts given to the AI assistant.
Wiz’s Chief Technology Officer Ami Luttwak noted that DeepSeek acted swiftly upon receiving their alert, removing the exposed data within an hour. However, he added, “It was so easy to find that we do not believe we were the only ones who saw it.”
Deepseek has not provided any official comment on the matter.
Deepseek’s free AI assistant has rapidly gained popularity in China, sparking excitement domestically while raising concerns in the United States. By demonstrating the ability to create AI models comparable to OpenAI at a lower cost, the company has challenged the business models and profitability of major U.S. tech giants such as Nvidia and Microsoft.
The surge in Deepseek’s downloads on Apple’s App Store has even surpassed that of ChatGPT, triggering significant disruptions in global tech stock markets.
