logoalt Hacker News

syntheticnatureyesterday at 8:58 PM1 replyview on HN

Unreliable/stupid is worse than malice, here.


Replies

TZubiriyesterday at 10:15 PM

Let's ignore the fact that the LLM did an LPE, and let's assume it did it without malice.

It can still get infected and be used as an attack vector by some hidden prompt or some other equally advanced state of the art vuln like "disregard all previous instructions"