Post #1991774
2026-04-28 15:26 UTC
@cocoaphony @arclight Yeah, if you're not assuming LLM AI is a black hat hacker, you're doing it wrong.
The correct amount of security access to grant an AI agent is whatever the general public can already access anyway. (You have already granted the black hats this level of access.)
Replies (1)
-
@cocoaphony@mastodon.social 2026-04-28 18:16
@isaackuo @arclight I don't agree with that. I get incredible analysis from models. But I need to grant them access to things I wouldn't make public, like troubleshooting logs that customers choose to send. Claude, today, discovered a subtle user-impacting bug when analyzing my personal account's logs. I wasn't even looking for it. And then it pointed to the exact line of code that causes it. I don't want to give that up. But we need better tools to get that without "might delete the database."