cm0002@infosec.pub to Funny@sh.itjust.works · 7 days agoWatching AI companies complain that their intellectual property is not being respected will never stop being hilariousinfosec.pubimagemessage-square2linkfedilinkarrow-up17arrow-down10cross-posted to: [email protected]
arrow-up17arrow-down1imageWatching AI companies complain that their intellectual property is not being respected will never stop being hilariousinfosec.pubcm0002@infosec.pub to Funny@sh.itjust.works · 7 days agomessage-square2linkfedilinkcross-posted to: [email protected]
minus-squareTar_Alcaran@sh.itjust.workslinkfedilinkarrow-up1·7 days agoAlso pictured here: Anthropic stating out loud their models will just give out all the “secret” and “secured” internal data to anyone who asks. Of course, that’s by design. LLMs can’t have any barrier between data and instructions, so they can never be secure.
Also pictured here: Anthropic stating out loud their models will just give out all the “secret” and “secured” internal data to anyone who asks.
Of course, that’s by design. LLMs can’t have any barrier between data and instructions, so they can never be secure.