🔴 NEW: Breaking LLMs: How Data Poisoning Attacks Work

Data poisoning lets attackers corrupt your AI model before it ever reaches your servers. Learn how backdoor attacks work, real 2024 incidents, and how to defend your training pipeline.

0:00 Intro
0:0

https://www.youtube.com/watch?v=LfJZxc8S6fA

#AISecurity #DataPoisoning #LLMAttacks #MachineLearningSecurity #Cybersecurity2024 #datapoisoningattackLLM #AIbackdoorattack #machinelearningsecurity

My favorite slide from a recent (?) talk by @simon

#LLMs #ai #ml #machinelearning #artificialintelligence #itsec #itsecurity #LLMattacks