Do you enjoy spending quality time threat modeling? That's fantastic! But, don't forget... "All models are wrong, but some are useful." - George Box Ask yoursel
The discussion centers on the practice of threat modeling, particularly in the context of AI and LLMs. It emphasizes the importance of collaborative threat modeling, suggesting that individuals should frequently engage with others, such as developers and testers, to gain diverse feedback. The core idea is that while models can be useful, they are inherently imperfect, highlighting the need for shared insights and regular communication in security practices. AI
IMPACT Encourages collaborative security practices for AI systems, suggesting a need for shared feedback loops.