Secure By Dezign
Model Inversion Attacks: How Adversaries Extract Your Training Data from LLMs
March 4 · 15 min · Season 1 · Episode 19 · 7.9 MB
0:00-15:57
Streams straight from the publisher. podnod never proxies or re-hosts episode audio.
When your AI becomes an unwitting data exfiltration tool. Technical breakdown of model inversion, membership inference, and training data extraction attacks with defensive countermeasures.
No links were found in this episode’s notes.