Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

What about this "newfangled AI"? Going into mindless panic over every big security team self-marketing post is not a rational behaviour even with AI.
 help



It's relentless, that's the thing. No human has the persistence to exploit this but you tell Claude Code to do it and come back 3 days and $100000 later and it's done.

It can't be reasoned with, it can't be bargained with, and it won't stop, ever, until your site is pwned!

Cool, but still: this has a massive list of ifs and buts to be exploitable that persistence won't fix.

Yes, it needs patching but it's hardly php2shell.

Also last I checked these agents are controlled by people - when's the last time US put anyone to jail for cyber attacks or at least fined them?


Actually this is very analogous to php2shell. That was also extremely difficult to exploit, yet AI got access to a live website within 2 hours and stole their bitcoin.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: