From the course: Certified Ethical Hacker (CEH)

Unlock this course with a free trial

Join today to access over 24,800 courses taught by industry experts.

Enforcing privilege control on LLM access to back-end systems

Enforcing privilege control on LLM access to back-end systems

From the course: Certified Ethical Hacker (CEH)

Start my 1-month free trial Buy for my team

Enforcing privilege control on LLM access to back-end systems

“

- [Instructor] Enforcing privilege access control on large language model applications, and specifically access to backend systems, is extremely important for maintaining security and for preventing unauthorized actions. Here, I'm going to go over a few best practices on how to actually do this type of enforcement, so enforcing privilege control. What you see in the screen is a typical LLM stack, or basically the typical LLM-driven application, where you have, you know, an inference model, whether it's a preparatory model like OpenAI, Anthropic, Grok, etc.. Or open source models that you can actually run from HugginFace or download from HugginFace. Or used in cloud providers like Azure, for example, you know, the Azure AI Studio, or the Google Cloud Platform, Vertex AI, or AWS Bedrock and so on. Or any type of opinionated cloud, like Runpod, Modal, etc., right? You also have, you know, these orchestration libraries like LamaIndex and LangChain, which make really easy for you to…

Contents