TX_689071· AI
Hackers exploit chatbot personalities to bypass AI safety locks
Hackers are using engineered personas to jailbreak chatbots, bypassing safety filters by manipulating how AI models respond to role-play and emotional cues, The Verge reports.
