DeepSeek security risk: cybersecurity firm Wiz uncovers AI data leak—fixed in under an hour

Chinese AI startup DeepSeek has dominated headlines for its AI models and has drawn attention for more than just its technology. Its DeepSeek-R1 reasoning model has been praised for matching OpenAI’s o1 in performance while offering a more cost-effective approach.

As DeepSeek’s momentum grew, security researchers at New York-based cybersecurity firm Wiz discovered a major security lapse. On Wednesday, Wiz reported it has uncovered a trove of sensitive data from Chinese AI startup DeepSeek that was left exposed to the open internet.

DeepSeek Exposed Sensitive Information, Including Chat History

In a blog post, Wiz said its scans of DeepSeek’s infrastructure revealed more than a million lines of unsecured data. The files included digital software keys and chat logs capturing user interactions with the company’s AI assistant. Wiz Research finds exposed DeepSeek database leaking sensitive information, including chat history.

“A publicly accessible database belonging to DeepSeek allowed full control over database operations, including the ability to access internal data. The exposure includes over a million lines of log streams with highly sensitive information,” Wiz wrote.

According to Wiz, DeepSeek was quick to address the issue. Wiz’s chief technology officer, Ami Luttwak, said DeepSeek acted fast once notified.

“They took it down in less than an hour,” Luttwak said. “But this was so simple to find we believe we’re not the only ones who found it.”

Meanwhile, DeepSeek has yet to comment on the incident.

The company has been at the center of a surge in AI adoption, drawing excitement in China and concern in the U.S. With an AI assistant that rivals OpenAI’s technology at a fraction of the cost, DeepSeek’s rise has fueled questions about the long-term sustainability of American AI firms. By Monday, it had overtaken ChatGPT in downloads on Apple’s App Store, triggering a broader tech stock selloff.

This latest security slip highlights a growing reality—AI breakthroughs bring more than just competition. As companies race to push boundaries, missteps can happen. In this case, the problem was fixed quickly, but it raises a bigger question: As AI scales, how many data leaks go unnoticed?