What’s prompt hacking ? | S03 E016 | Build On Generative AI

You've probably heard about prompt engineering, a way to craft your prompt so it influences the output of an LLM. Here let's take prompt engineering to the next level and explore the world of prompt hacking. Prompt hacking is a term used to describe attacks that exploit vulnerabilities of LLMs, by manipulating their inputs or prompts. We'll see different prompt hacking techniques such as Prompt injection, prompt leaking and jailbreaking. We'll also see defensive measures such as filtering, sandwich defense, XML tagging, random sequence enclosure... Finally we'll see how you can easily implement those defensive measures with Amazon Guardrails.

Check out the recording here:

Feedback:
Did you like this episode ? What are the other topics you would like us to talk about ? Let us know HERE.

Shared links:

Reach out to the hosts: