SicurezzaENFreitag, 28. November 2025

Prompt Injection Through Poetry

Schneier on Security

Fonte Esterna

Riepilogo

In a new paper, “Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models,” researchers found that turning LLM prompts into poetry resulted in jailbreaking the models...

Articolo Esterno

Questo articolo proviene da Schneier on Security ed è ospitato lì. Linkiamo solo a fonti esterne.