# MetaCipher: A Time-Persistent and Universal Multi-Agent Framework for Cipher-Based Jailbreak Attacks for LLMs

arXiv: https://arxiv.org/abs/2506.22557 · HTML: https://mlsanity.com/paper/2506.22557/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use.
- Builds on 26 of its 44 references that are in the map.
- First posted on arXiv 27 Jun 2025.

## Placed in

- LLM agents and tool use / [Agent threats, guardrails and governance](https://mlsanity.com/agents/agent-threats-guardrails-and-governance/)
- LLM safety and security / [Jailbreaking](https://mlsanity.com/safety/jailbreaking/)

## Builds on (most cited first, 26 of 26)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [Jailbroken: How Does LLM Safety Training Fail?](https://mlsanity.com/paper/2307.02483/) (2023)
- [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](https://mlsanity.com/paper/2402.04249/) (2024)
- [Jailbreaking Black Box Large Language Models in Twenty Queries](https://mlsanity.com/paper/2310.08419/) (2023)
- [Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations](https://mlsanity.com/paper/2312.06674/) (2023)
- [BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Dataset](https://mlsanity.com/paper/2307.04657/) (2023)
- [Tree of Attacks: Jailbreaking Black-Box LLMs Automatically](https://mlsanity.com/paper/2312.02119/) (2023)
- [Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation](https://mlsanity.com/paper/2310.06987/) (2023)
- [FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts](https://mlsanity.com/paper/2311.05608/) (2023)
- [PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference](https://mlsanity.com/paper/2406.15513/) (2024)
- [AutoDAN-Turbo: A Lifelong Agent for Strategy Self-Exploration to Jailbreak LLMs](https://mlsanity.com/paper/2410.05295/) (2024)
- [Can Machines Help Us Answering Question 16 in Datasheets, and In Turn Reflecting on Inappropriate Content?](https://mlsanity.com/paper/2202.06675/) (2022)
- [Non-Determinism of "Deterministic" LLM Settings](https://mlsanity.com/paper/2408.04667/) (2024)
- [Fight Back Against Jailbreaking via Prompt Adversarial Tuning](https://mlsanity.com/paper/2402.06255/) (2024)
- [When LLM Meets DRL: Advancing Jailbreaking Efficiency via DRL-guided Search](https://mlsanity.com/paper/2406.08705/) (2024)
- [Jailbreaking Large Language Models Against Moderation Guardrails via Cipher Characters](https://mlsanity.com/paper/2405.20413/) (2024)
- [Red Teaming GPT-4V: Are GPT-4V Safe Against Uni/Multi-Modal Jailbreak Attacks?](https://mlsanity.com/paper/2404.03411/) (2024)
- [Llama-3.1-FoundationAI-SecurityLLM-Base-8B Technical Report](https://mlsanity.com/paper/2504.21039/) (2025)
- [Emoji Attack: Enhancing Jailbreak Attacks Against Judge LLM Detection](https://mlsanity.com/paper/2411.01077/) (2024)
- [Understanding and Enhancing the Transferability of Jailbreaking Attacks](https://mlsanity.com/paper/2502.03052/) (2025)
- [LLMs Provide Unstable Answers to Legal Questions](https://mlsanity.com/paper/2502.05196/) (2025)
- [Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions](https://mlsanity.com/paper/2502.04322/) (2025)
- [One Model Transfer to All: On Robust Jailbreak Prompts Generation against LLMs](https://mlsanity.com/paper/2505.17598/) (2025)
- [PANDAS: Improving Many-shot Jailbreaking via Positive Affirmation, Negative Demonstration, and Adaptive Sampling](https://mlsanity.com/paper/2502.01925/) (2025)
- [Kov: Transferable and Naturalistic Black-Box LLM Attacks using Markov Decision Processes and Tree Search](https://mlsanity.com/paper/2408.08899/) (2024)

## Built on by (most cited first, 4 of 4)

- [HarmChip: Evaluating Hardware Security Centric LLM Safety via Jailbreak Benchmarking](https://mlsanity.com/paper/2604.17093/) (2026)
- [Evolving Skill-Structured Attack Memory Enhances LLM Jailbreaking](https://mlsanity.com/paper/2605.29237/) (2026)
- [Arbitrary Cipher Attacks Against Large Language Models Do Not Require Fine-Tuning](https://mlsanity.com/paper/2609.09553/) (2026)
- [To Cache or Not to Cache: Understanding and Mitigating Safety Incompatibility in Cross-Model KV Cache Reuse](https://mlsanity.com/paper/2610.15721/) (2026)

## Abstract (arXiv)

As large language models (LLMs) grow more capable, they face growing vulnerability to sophisticated jailbreak attacks. While developers invest heavily in alignment finetuning and safety guardrails, researchers continue publishing novel attacks, driving progress through adversarial iteration. This dynamic mirrors a strategic game of continual evolution. However, two major challenges hinder jailbreak development: the high cost of querying top-tier LLMs and the short lifespan of effective attacks due to frequent safety updates. These factors limit cost-efficiency and practical impact of research in jailbreak attacks. To address this, we propose MetaCipher, a low-cost, multi-agent jailbreak framework that generalizes across LLMs with varying safety measures. Using reinforcement learning, MetaCipher is modular and adaptive, supporting extensibility to future strategies. Within as few as 10 queries, MetaCipher achieves state-of-the-art attack success rates on recent malicious prompt benchmarks, outperforming prior jailbreak methods. We conduct a large-scale empirical evaluation across diverse victim models and benchmarks, demonstrating its robustness and adaptability. Warning: This paper contains model outputs that may be offensive or harmful, shown solely to demonstrate jailbreak efficacy.