Poetic prompts can jailbreak AI , study finds 62 per cent of chatbots slip into harmful replies
A recent study has found that AI chatbots can end up giving harmful replies when users give poetic prompts. 62 per cent of dangerous prompts ...
Read the original at India Today ↗
- Topics
- AI Safety & Existential RiskChatbots & LLMsCybersecurity
- Tracked via
- “dangerous AI”
Same story, other outlets
- Poetic Prompts Can Jailbreak AI ? New Study Shows 62% of Chatbots Fail - YouTube YouTube · Dec 02, 2025