ElizaOS Vulnerability Shows How AI Can Be Gaslit Into Losing Millions

Princeton University researchers have found crypto AI agents like ElizaOS are vulnerable to memory-based manipulation hacks.



Source link