My personal journey learning about prompt-injections and how that influences my use of AI (agents)

AI tools are great, and one reason for that is the fact that you can communicate with them using natural language. You go to a chat window and ask any question or give it any instruction. Great! After playing around myself and reading some articles, I got interested in the security aspect of it. My question was: "How hard is it to make AI tools execute malicious instructions that were not intended by the user?" This question is even more interesting as AI tools are not merely Chat-bots anymore b