Pith. sign in

REVIEW 15 cited by

Open Sesame! Universal Black Box Jailbreaking of Large Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2309.01446 v4 pith:OTA2I4NJ submitted 2023-09-04 cs.CL cs.CVcs.NE

Open Sesame! Universal Black Box Jailbreaking of Large Language Models

classification cs.CL cs.CVcs.NE
keywords alignmentllmsmodeluniversalapproachattackblackintent
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Large language models (LLMs), designed to provide helpful and safe responses, often rely on alignment techniques to align with user intent and social guidelines. Unfortunately, this alignment can be exploited by malicious actors seeking to manipulate an LLM's outputs for unintended purposes. In this paper we introduce a novel approach that employs a genetic algorithm (GA) to manipulate LLMs when model architecture and parameters are inaccessible. The GA attack works by optimizing a universal adversarial prompt that -- when combined with a user's query -- disrupts the attacked model's alignment, resulting in unintended and potentially harmful outputs. Our novel approach systematically reveals a model's limitations and vulnerabilities by uncovering instances where its responses deviate from expected behavior. Through extensive experiments we demonstrate the efficacy of our technique, thus contributing to the ongoing discussion on responsible AI development by providing a diagnostic tool for evaluating and enhancing alignment of LLMs with human intent. To our knowledge this is the first automated universal black box jailbreak attack.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 15 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. The Salami Slicing Threat: Exploiting Cumulative Risks in LLM Systems

    cs.CR 2026-04 unverdicted novelty 6.0

    Salami Attack chains low-risk inputs to cumulatively trigger high-risk LLM behaviors, achieving over 90% success on GPT-4o and Gemini while resisting some defenses.

  2. Stop Tracking Me! Proactive Defense Against Attribute Inference Attack in LLMs

    cs.CR 2026-02 conditional novelty 6.0

    TRACE-RPS drops LLM attribute inference accuracy from around 50% to below 5% via fine-grained anonymization plus a two-stage rejection optimization.

  3. Evolve the Method, Not the Prompts: Evolutionary Synthesis of Jailbreak Attacks on LLMs

    cs.CL 2025-11 unverdicted novelty 6.0

    EvoSynth evolves code-based jailbreak algorithms via multi-agent self-correction, reaching 85.5% ASR on Claude-Sonnet-4.5 and 95.9% average across targets with greater diversity.

  4. SoK: Systematizing LLM Prompt Security: Taxonomies, Datasets, and Unified Evaluation of Attacks and Defenses

    cs.CR 2025-10 conditional novelty 6.0

    A systemization of LLM jailbreak security that adds linked taxonomies, an evaluation platform, and JailbreakDB, while its main attack–defense comparison results remain deferred.

  5. Toward Principled LLM Safety Testing: Solving the Jailbreak Oracle Problem

    cs.CR 2025-06 unverdicted novelty 6.0

    Formalizes the jailbreak oracle problem for LLMs and introduces Boa, a two-phase breadth-first then depth-first search system to solve it efficiently.

  6. Exploring the Secondary Risks of Large Language Models

    cs.LG 2025-06 unverdicted novelty 6.0

    Introduces secondary risks as a new class of LLM failures from benign prompts, defines two primitives, proposes SecLens search framework, and releases SecRiskBench showing risks are widespread across 16 models.

  7. Towards an AI co-scientist

    cs.AI 2025-02 unverdicted novelty 6.0

    A multi-agent AI system generates novel biomedical hypotheses that show promising experimental validation in drug repurposing for leukemia, new targets for liver fibrosis, and a bacterial gene transfer mechanism.

  8. JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models

    cs.CR 2024-03 accept novelty 6.0

    JailbreakBench supplies an evolving set of jailbreak prompts, a 100-behavior dataset aligned with usage policies, a standardized evaluation framework, and a leaderboard to enable comparable assessments of attacks and ...

  9. A StrongREJECT for Empty Jailbreaks

    cs.LG 2024-02 conditional novelty 6.0

    StrongREJECT provides a standardized benchmark and evaluator for jailbreak attacks that aligns better with human judgments than prior methods and reveals that successful jailbreaks often reduce model capabilities.

  10. Low-Resource Languages Jailbreak GPT-4

    cs.CL 2023-10 conditional novelty 6.0

    Translating unsafe inputs to low-resource languages jailbreaks GPT-4 at rates on par with or exceeding state-of-the-art attacks.

  11. Certifiable Safe RLHF: Semantic Grounding and Fixed Penalty Constraint Optimization for Safer LLM Alignment

    cs.LG 2025-10 reject novelty 5.0

    A fixed ReLU penalty and a semantically labeled cost model are proposed to make RLHF safer, but the 'certifiable' guarantee is not fully supported.

  12. TrustLLM: Trustworthiness in Large Language Models

    cs.CL 2024-01 unverdicted novelty 5.0

    TrustLLM defines eight trustworthiness principles, creates a six-dimension benchmark, and evaluates 16 LLMs showing proprietary models generally lead but some open-source ones are close while over-calibration can hurt...

  13. From Concept-Aligned Tokens to Vulnerable Features: Mechanistic Localization of Jailbreaks

    cs.CL 2026-04 unverdicted novelty 4.0

    Jailbreak success in Gemma-2-2B is driven by identifiable feature subgroups in mid-to-later layers rather than prompts alone, as shown by consistent steering effects across three grouping methods.

  14. From Concept-Aligned Tokens to Vulnerable Features: Mechanistic Localization of Jailbreaks

    cs.CL 2026-04 conditional novelty 4.0

    Amplifying sparse autoencoder feature subgroups in mid-to-late layers of Gemma-2-2B increases harmfulness of responses, with single-token-driven grouping matching cluster-based grouping.

  15. Jailbreak Attacks and Defenses Against Large Language Models: A Survey

    cs.CR 2024-07 accept novelty 4.0

    A survey that creates taxonomies for jailbreak attacks and defenses on LLMs, subdivides them into sub-classes, and compares evaluation approaches.