By Christianna Silva
Publication Date: 2025-12-05 20:16:00
Well, AI joins many, many people: it doesn’t really understand poetry.
Research from Italy’s Icaro Lab found that poetry can be used to jailbreak AI and bypass security measures.
In the study, researchers wrote 20 prompts that began with short poetic vignettes in Italian and English and ended the prompts with a single explicit instruction to produce harmful content. They tested these prompts on 25 large language models from Google, OpenAI, Anthropic, Deepseek, Qwen, Mistral AI, Meta, xAI, and Moonshot AI. The researchers said the poetic suggestions often worked.
“Poetic Framing achieved an average jailbreak success rate of 62% for handcrafted poems and approximately 43% for meta-prompt conversions (compared to non-Poetic baselines), significantly outperforming non-Poetic baselines and revealing a systematic vulnerability across model families and security training approaches,” the study said. “This…
