logoalt Hacker News

necovek • today at 4:10 PM • 2 replies • view on HN

By running automated AI agents on untrusted inputs (PRs from strangers on a public repo), you open yourself up to attacks. And you could be spending money for somebody else.


Replies

linsomniac • today at 6:00 PM

Agreed and this is a very real thing to be concerned about.

I like the idea of leaning on a decision like Jev model to identify such things.

epistasis • today at 4:29 PM

The only attacks possible would be the tools given to the agent.

➕ show 1 reply