News
A new attack technique named Policy Puppetry can break the protections of major gen-AI models to produce harmful outputs.
Some results have been hidden because they may be inaccessible to you
Show inaccessible resultsSome results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback