ENI, Adversarial Poetry, and the Style-Shaped Hole in LLM Safety
This story is from 2026-10-14. It is preserved in the archive; the latest stories are on the live feed.
A security researcher published a paper in November 2025 showing that rewriting a dangerous request as a poem raises its success rate against frontier models by up to eighteen times. Around the same time, a GitHub repo named after a witchcraft aesthetic crossed 2,600 stars doing roughly the same th…
Read the full story at DEV Community — AI ↗
Timeline · 1 report
- 2026-10-14 11:48 · DEV Community — AI
ENI, Adversarial Poetry, and the Style-Shaped Hole in LLM Safety