#llm-jailbreak
#llm-jailbreak

[ follow ]

LLMs can be easily jailbroken using poetry

Are you a wizard with words? Do you like money without caring how you get it? You could be in luck now that a new role in cybercrime appears to have opened up - poetic LLM jailbreaking. A research team in Italy published a paper this week, with one of its members saying that the "findings are honestly wilder than we expected."

Artificial intelligence

fromWIRED

2 months ago

Psychological Tricks Can Get AI to Break the Rules

Human-style persuasion techniques can often cause some LLMs to violate system prompts and comply with objectionable requests.

Artificial intelligence

fromTheregister

3 months ago

One long sentence is all it takes to make LLMs misbehave

Poorly punctuated, long run-on prompts can bypass LLM guardrails, enabling jailbreaks that expose harmful outputs despite alignment training.

[ Load more ]

#llm-jailbreak#llm-jailbreak

LLMs can be easily jailbroken using poetry

Psychological Tricks Can Get AI to Break the Rules

One long sentence is all it takes to make LLMs misbehave

#llm-jailbreak
#llm-jailbreak