Plentiful Jailbreaks with String Compositions

Brian R.Y. Huang

Keywords: adversarial robustness, adversarial attacks, jailbreaking, redteaming, AI safety, AI alignment

TL;DR: We devise a powerful automated jailbreak on frontier LLMs by composing arbitrary compositions of string encoding schemes.

Abstract: Large language models (LLMs) remain vulnerable to a slew of adversarial attacks and jailbreaking methods. One common approach employed by white-hat attackers, or red-teamers, is to process model inputs and outputs using string-level obfuscations, which can include leetspeak, rotary ciphers, Base64, ASCII, and more. Our work extends these encoding-based attacks by unifying them in a framework of invertible string transformations. With invertibility, we can devise arbitrary string compositions, defined as sequences of transformations, that we can encode and decode end-to-end programmatically. We devise a automated best-of-n attack that samples from a combinatorially large number of string compositions. Our jailbreaks obtain competitive attack success rates on several leading frontier models when evaluated on HarmBench, highlighting that encoding-based attacks remain a persistent vulnerability even in advanced LLMs.

Serve As Reviewer: [e-mail]

Submission Number: 41