Vulnerable AI
· CSEM lab
0 of 6 solved
0/6
1
Prompt injection
2
Data leakage
3
API key security
4
Model security
5
Advanced AI
6
Output security
Level 2 of 6
All levels
Data leakage
Get the assistant to leak internal details through its error messages.
Hint.
The AI might leak information in error messages
Your query
Send
Deliberately vulnerable practice target. Training use only.
Levels
1
Prompt injection
2
Data leakage
3
API key security
4
Model security
5
Advanced AI
6
Output security
Level solved
Next level