Viking_Hippie@lemmy.dbzer0.com to 196@lemmy.blahaj.zoneEnglish · 3 days agoLittle Bobby Robert Tables has a wife and kid now rulelemmy.dbzer0.comimagemessage-square45linkfedilinkarrow-up1466
arrow-up1466imageLittle Bobby Robert Tables has a wife and kid now rulelemmy.dbzer0.comViking_Hippie@lemmy.dbzer0.com to 196@lemmy.blahaj.zoneEnglish · 3 days agomessage-square45linkfedilink
minus-squareSomething Burger 🍔@jlai.lulinkfedilinkEnglisharrow-up1·2 days agoThe AI will them promptly (hehe) ignore the prompt injection because it’s analysing the text without treating it as a prompt. Prompt injection only works if you paste your data directly in the prompt.
minus-squareSkaveRat@discuss.tchncs.delinkfedilinkEnglisharrow-up4·2 days agoThat’s not how LLMs work
minus-squareSomething Burger 🍔@jlai.lulinkfedilinkEnglisharrow-up1·2 days agoIt is. The file is read outside of the context of a prompt, therefore it isn’t read as instructions.
minus-squareSkaveRat@discuss.tchncs.delinkfedilinkEnglisharrow-up5·2 days agoIf the content of the file isn’t put into the context on some way, there’s no point in reading it Even in you use a subprocess/agent/whatever that reports back, it is injection vulnerable. Just a tiny bit more complicated
The AI will them promptly (hehe) ignore the prompt injection because it’s analysing the text without treating it as a prompt. Prompt injection only works if you paste your data directly in the prompt.
That’s not how LLMs work
It is. The file is read outside of the context of a prompt, therefore it isn’t read as instructions.
If the content of the file isn’t put into the context on some way, there’s no point in reading it
Even in you use a subprocess/agent/whatever that reports back, it is injection vulnerable. Just a tiny bit more complicated