Wiz Research, a New York-based cybersecurity company, has reported that it has discovered a huge amount of confidential data from Chinese artificial intelligence startup DeepSeek. They were made publicly available on the Internet.
In a blog post, Wiz cyber experts said that a scan of DeepSeek's infrastructure revealed that the company had accidentally left more than a million rows of data unprotected. Among them were digital software keys and chats that probably recorded requests from users to the company's free AI assistant.
Wiz's CTO said that the analysts contacted the DeepSeek team, and the company promptly secured the data.
“They took it down in less than an hour,” Ami Luttwak said. “But this was so simple to find that we believe we’re not the only ones who found it.”
By the way, this is not the first scandal related to DeepSeek. Recently, OpenAI, the developer of ChatGPT, reported that it had evidence that a Chinese AI lab used its model to train its own artificial intelligence system.