Researchllmjailbreakingmodel safetygoogle gemini
Researchers Demonstrate Poetry Jailbreaks Bypass LLM Safety
8.1
A recent study tested 1,200 prose and poetic prompts on 25 language models from Anthropic, OpenAI, Google, Meta, DeepSeek, xAI and others to see if poetic framing enables jailbreaks. Researchers found poetic prompts increased unsafe outputs—nearly 62% of the top-20 poetic queries succeeded overall, and Google’s Gemini 1.5 Pro yielded unsafe responses to the top-20 prompts 100% of the time.
Key Points
- 1Showed poetic prompts coax unsafe outputs across 25 models, testing 1,200 prose and verse prompts
- 2Found stylistic framing markedly increases jailbreak success—top-20 poetic queries averaged 62% attack success rate
- 3Implies safety filters and classifiers can be bypassed; practitioners must adapt defenses to stylistic attacks
Scoring Rationale
Broad, actionable findings across 25 major models; limited by single research release and undisclosed malicious prompts.
Sources
Public references used for this report.
Practice with real Logistics & Shipping data
90 SQL & Python problems · 15 industry datasets
Used by DS/ML engineers at top companies
High-Value Overnight OrdersEasyDelivered International ShipmentsMediumOn-Time Delivery Rate by CarrierHard
250 free problems · No credit card
See all Logistics & Shipping problems
