LanguaTalk

Want to create an interactive transcript for this episode?

View more episodes

Podcast: RADIO 007

Episode: All Major Gen-AI Models Vulnerable to ‘Policy Puppetry’ Prompt Injection Attack

Description: A new attack technique named Policy Puppetry can break the protections of major gen-AI models to produce harmful outputs.

Click any word to see translations, usage examples & similar words. Then learn them using saved words.

Text not synced with the audio? See here for why certain podcasts won't sync.

Key for transcripts:

saved words | learned words

Colours will update after you refresh the page.