The TIP of the Iceberg: Revealing a Hidden Class of Task-in-Prompt Adversarial Attacks on LLMs

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We present a novel class of jailbreak adversarial attacks on LLMs, termed Task-in-Prompt (TIP) attacks. Our approach embeds sequence-to-sequence tasks (e.g., cipher decoding, riddles, code execution) into the model's prompt to indirectly generate prohibited inputs. To systematically assess the effectiveness of these attacks, we introduce the PHRYGE benchmark. We demonstrate that our techniques successfully circumvent safeguards in six state-of-the-art language models, including GPT-4o and LLaMA 3.2. Our findings highlight critical weaknesses in current LLM safety alignment and underscore the urgent need for more sophisticated defence strategies.

Original languageEnglish
Title of host publicationLong Papers
EditorsWanxiang Che, Joyce Nabende, Ekaterina Shutova, Mohammad Taher Pilehvar
PublisherAssociation for Computational Linguistics (ACL)
Pages6716-6730
Number of pages15
ISBN (Electronic)9798891762510
DOIs
Publication statusPublished - 1 Jan 2025
Event63rd Annual Meeting of the Association for Computational Linguistics, ACL 2025 - Vienna, Austria
Duration: 27 Jul 20251 Aug 2025

Publication series

NameProceedings of the Annual Meeting of the Association for Computational Linguistics
Volume1
ISSN (Print)0736-587X

Conference

Conference63rd Annual Meeting of the Association for Computational Linguistics, ACL 2025
Country/TerritoryAustria
CityVienna
Period27/07/251/08/25

Fingerprint

Dive into the research topics of 'The TIP of the Iceberg: Revealing a Hidden Class of Task-in-Prompt Adversarial Attacks on LLMs'. Together they form a unique fingerprint.

Cite this