By using this website, you agree to our Privacy Policy and Terms of Use.
Accept
Craftium.AICraftium.AICraftium.AI
  • Home
  • News
  • Catalog
  • Collections
  • Blog
Font ResizerAa
Craftium.AICraftium.AI
Font ResizerAa
Пошук
  • Home
  • News
  • Catalog
  • Collections
  • Blog
Follow US
  • Terms of Use
  • Privacy Policy
  • Copyright
  • Feedback
© 2024-2025 Craftium.AI.

DeepSeek’s Open Database Threatened User Security

Wiz discovered a million records in the database, including chats and API keys, with no authentication required

Alex Dubenko
Alex Dubenko
Published: 31.01.2025
News
DeepSeek
Illustrative image
SHARE

Researchers from Wiz reported the discovery of an open database belonging to the Chinese startup DeepSeek, which contained sensitive information. The open access to the database was found within minutes, with no authentication required. Among the exposed data were users’ chat histories, API authentication keys, and system logs.

The DeepSeek database was hosted on the open-source data management system ClickHouse and contained over a million log entries. Wiz noted that this situation could have led to full control over the database and potential privilege escalation within the DeepSeek environment. This could have given attackers access to the startup’s internal systems.

Read also

Learning with AI
ChatGPT received Study Mode for user learning mode
Microsoft prepares its Copilot service for the launch of GPT-5
ChatGPT will independently choose the model for each user request

After being notified by Wiz, DeepSeek promptly secured the database. There is no information on whether anyone else accessed the exposed data, but researchers noted that it would not be surprising given how easy it was to find the database.

Interestingly, DeepSeek’s systems have a design similar to OpenAI’s, down to the API key format. Earlier this week, OpenAI accused DeepSeek of using its data to train its AI models.

Google DeepMind and OpenAI models won gold at the mathematics olympiad
GPT-5 is being tested for science and complex programming tasks
OpenAI urges caution when using the ChatGPT agent
OpenAI introduced ChatGPT Agent, allowing AI to delegate complex tasks
ChatGPT to Receive Agent Mode for Working with Files and Reports
TAGGED:DeepSeekOpenAISecurity
Leave a Comment

Leave a Reply Cancel reply

Follow us

XFollow
YoutubeSubscribe
TelegramFollow
MediumFollow

Popular News

Mistral AI
Mistral offers the open language model Voxtral for speech
17.07.2025
Grok 4
Grok 4 from xAI promises to become the new standard for developers
06.07.2025
Siri
Apple Tests OpenAI and Anthropic Models for Next-Generation Siri
01.07.2025
AI machine
The number of AI users reaches 1.8 billion, only 3% use it for a fee
01.07.2025
Chatbots in messengers
Meta chatbots will gain the ability to independently send messages to users
04.07.2025

Читайте також

Monitoring AI Thoughts
News

Researchers Urge to Maintain Transparency of Thought in AI Models

16.07.2025
OpenAI
News

OpenAI prepares to launch its own browser with built-in AI

11.07.2025
Spies in AI
News

OpenAI strengthens protection against competitor espionage

09.07.2025

Craftium AI is a team that closely follows the development of generative AI, applies it in their creative work, and eagerly shares their own discoveries.

Navigation

  • News
  • Reviews
  • Collections
  • Blog

Useful

  • Terms of Use
  • Privacy Policy
  • Copyright
  • Feedback

Subscribe for AI news, tips, and guides to ignite creativity and enhance productivity.

By subscribing, you accept our Privacy Policy and Terms of Use.

Craftium.AICraftium.AI
Follow US
© 2024-2025 Craftium.AI
Subscribe
Level Up with AI!
Get inspired with impactful news, smart tips and creative guides delivered directly to your inbox.

By subscribing, you accept our Privacy Policy and Terms of Use.

Welcome Back!

Sign in to your account

Username or Email Address
Password

Lost your password?