3 ms·

I'm as paranoid about this as you, but this type of verification seems easier today with AI tools. I'm not aware of any that do this, but if LLMs can give insig

by imiric 3y ago

I'm as paranoid about this as you, but this type of verification seems easier today with AI tools. I'm not aware of any that do this, but if LLMs can give insight about what a piece of code is doing, they can surely be trained to detect possible suspicious behavior. Perhaps even by inspecting a binary, but certainly by processing code.

BowBun 3y ago
I think that's a great use-case. I'd love a real-time security scanning system covering as many open source projects out there as possible.
krpl 3y ago
https://any.run/cybersecurity-blog/chatgpt-powered-analysis-method/ https://any.run/cybersecurity-blog/chatgpt-powered-analysis-... Sort of did this
lxgr 3y ago
Maybe for well-intended code (and even there I have my doubts – the halting problem says hi!), but most definitely not for malicious backdoors at this point.