In a recent cybersecurity revelation, New York-based cybersecurity firm Wiz said that it recently uncovered an inadvertently exposed database of DeepSeek, a prominent Chinese artificial intelligence (AI) company, leaking sensitive data, including chat messages and personal information of the users, to the open internet.
DeepSeek, founded in Hangzhou in 2023, has rapidly ascended in the AI industry due to its innovative chatbot models, particularly the DeepSeek-R1 reasoning model.
This AI model has been praised for its performance, rivaling U.S. counterparts like OpenAI’s o1 while utilizing fewer resources.
In a blog post published on Wednesday, Wiz said it identified a publicly accessible ClickHouse database related to DeepSeek.
The exposed database allowed complete control over database operations, including the ability to access internal data. It included over a million lines of log streams containing chat history, secret keys, backend details, and other highly sensitive information.
“Within minutes, we found a publicly accessible ClickHouse database linked to DeepSeek, completely open and unauthenticated, exposing sensitive data. It was hosted at oauth2callback.deepseek.com:9000 and dev.deepseek.com:9000,” the cybersecurity wrote in its blog post.
“This database contained a significant volume of chat history, backend data and sensitive information, including log streams, API Secrets, and operational details.
“More critically, the exposure allowed for full database control and potential privilege escalation within the DeepSeek environment, without any authentication or defense mechanism to the outside world.”
On discovering the exposed database, Wiz’s co-founder Ami Luttwak said that their research team immediately and responsibly disclosed the issue to DeepSeek, which then quickly secured the database.
“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.”
Although DeepSeek acted quickly to fix the issue, this discovery has raised significant concerns regarding data privacy and the potential for misuse by government entities.
DeepSeek has currently limited user registrations because of an ongoing cyberattack. Just 2 days back, another cybersecurity firm was able to jailbreak Deepseek.
These incidents underscores the growing challenges in ensuring data security and privacy, particularly with the rapid advancement of AI technologies.
DeepSeek has yet to comment on the issue.
