Exploring how far cyber safety approaches can lend a hand mitigate dangers in generative AI programs
From bugs to bypasses: adapting vulnerability disclosure for AI safeguards From bugs to bypasses: adapting vulnerability disclosure for AI safeguards