Learn how LLMs process text and craft effective prompts for security and adversarial testing, from TryHackMe’s new AI Security Path.

Artificial Intelligence is rapidly becoming part of EVERYTHING, including cybersecurity, from automation and detection to analysis and offensive security research.

As a result, understanding how AI systems work, including responding to prompts, is slowly becoming an important skill for cybersecurity professionals.

One of the core concepts behind interacting with AI systems is Prompt Engineering, the ability to structure prompts in ways that influence how AI responds, behaves, and interprets instructions.

Recently, TryHackMe introduced a new AI Security Path, and one of the rooms in the first section, AI Fundamentals, is Prompt Engineering.

In this walkthrough, I documented the steps taken to complete the tasks and the key takeaways from the room

Link to the Room: https://tryhackme.com/room/promptengineeringaisec

Task 2: LLM Fundamentals

1. What is the term for the smallest units that an LLM breaks text into in order to process it?

Answer: Tokens