How to Detect LLM’s Vulnerabilities: Giskard

[@SridharKumarKannam]

If you found this content useful, pleases consider sharing it with others who might benefit. Your support is greatly appreciated :)

[@thevarayan]

Excellent work

[@SridharKumarKannam]

Thanks for your support :)

[@Shhukoihee]

I have tried giskard sir but when i am starting giskard server [ giskard hub start] its started but not able to get free licence part

[@SridharKumarKannam]

Sorry, I've not used giskard-hub. Can't you use giskard like a python library in your code? Why you need the hub?

[@Shhukoihee]

@@SridharKumarKannam so I am cyber security professional and now working in AI Security
And one of my clients has an LLM model setup now I have to do the Vulnerability scan using giskard and for that I required giskard
Correct me if I am wrong and help me out if you know any other way like using only the python library How can I scan the client LLM model ?

[@SridharKumarKannam]

@@Shhukoihee you want to scan the input and output from llm or the llm model itself?

[@dhruvchugh9908]

Can you please share the source code?

[@SridharKumarKannam]

The link added in the description. By default they use OpenAI models, I've added the code to use Ollama, also in the description...