WASHINGTON, Jan 29 (Reuters) - New York-based cybersecurity firm Wiz reported discovering a cache of sensitive data inadvertently exposed online by the Chinese artificial intelligence startup DeepSeek.
According to Wiz, scans of DeepSeek's system revealed over a million unsecured data entries, including software keys and chat logs displaying user interactions with the company's AI assistant.
Wiz's chief technology officer, Ami Luttwak, mentioned that DeepSeek promptly secured the data within an hour of being informed by their firm. However, Luttwak expressed concern that the data's vulnerability may have been exploited by other parties beyond their detection.
At the time of reporting, DeepSeek had not responded to requests for comments.
DeepSeek's rapid rise in the AI sector has generated interest and raised questions about the sustainability of the business models and profit margins of prominent U.S. AI companies like Nvidia and Microsoft, as DeepSeek's technology capacitates comparable capabilities to OpenAI's at a significantly lower cost.
The surge in downloads of DeepSeek's AI assistant from Apple's App Store within a short span has surpassed its U.S. competitor ChatGPT, indicating a notable market impact.