← back
arXivSiyu Chen, Haoran Wang, Xiaojian Li, Yao Huang, Yinpeng Dong, Wei XuWed, Sep 2, 2026, 3:35 AM PDT
score 16.2

New attack method breaks AI safety by mimicking normal conversation over multiple turns

Original: Before the Script, Set the Stage: How Worldview Simulation Amplifies Psychologically Grounded Persuasion in Multi-Turn Jailbreaking

Source: arxiv.org

Writing ELI5 summary…