English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Most traditional AI safety research has approached AI models as machines andcentered on algorithm-focused attacks developed by security experts. As largelanguage models (LLMs) become increasingly common and competent, non-expertusers can also impose risks during daily interactions. This paper introduces anew perspective to jailbreak LLMs as human-like communicators, to explore thisoverlooked intersection between everyday language interaction and AI safety.Specifically, we study how to persuade LLMs to jailbreak them. First, wepropose a persuasion taxonomy derived from decades of social science research.Then, we apply the taxonomy to automatically generate interpretable persuasiveadversarial prompts (PAP) to jailbreak LLMs. Results show that persuasionsignificantly increases the jailbreak performance across all risk categories:PAP consistently achieves an attack success rate of over $92\%$ on Llama 2-7bChat, GPT-3.5, and GPT-4 in $10$ trials, surpassing recent algorithm-focusedattacks. On the defense side, we explore various mechanisms against PAP and,found a significant gap in existing defenses, and advocate for more fundamentalmitigation for highly interactive LLMs

How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMs