‘Signal’ President and VP warn agentic AI is insecure, unreliable, and a surveillance nightmare

TLDR

This talk at 39C3 discusses the dangers of integrating agentic AI into operating systems and applications, focusing on privacy and personal agency. It defines agentic AI, examines the threat it poses to privacy-first applications, and presents a framework for remediation, including empowering developers, granting granular user control, and mandating transparency.

16Score: 16

0 Comments