[CoffeeGoddes]

DeepSeek iOS App Raises Security Concerns

• Unencrypted Data: Transmits sensitive information without encryption.
  
• Weak Encryption: Uses outdated security methods.
  
• Data Sent to China: Linked to ByteDance servers.
  
• Security Risks: Exposes users to attacks and surveillance.
  
• Global Bans: Australia, Italy, and others have banned it.
  
• Ongoing Investigations: Ireland and France are probing its risks.
  
  

Should governments ban it or just warn users?
Or maybe they will silence everything like usual and keep the backdoors for themselfs?

[lunastr127]

It would be foolish to use Deepseek app itself while expecting privacy, on the other hand the same thing happens with ChatGPT and any other US-based app.

[LockedIN]

Seems like a lot of mostly smaller LLM apps/ websites in general have poor security. Heard of the omnigpt breach?

[4ng3rF1st]

Seems like a lot of mostly smaller LLM apps/ websites in general have poor security. Heard of the omnigpt breach?

That is very recent, but there are several other compromise to AI websites, especially thoose free img/art services around the web