# Dynamic Optimization and Safety Indicator Injection for Jailbreaking Text-to-Image Models with Multimodal Safety Filters

arXiv: https://arxiv.org/abs/2505.18979 · HTML: https://mlsanity.com/paper/2505.18979/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Vision-language and text-to-image safety, one of 12 directions in LLM safety and security.
- Builds on 29 of its 44 references that are in the map.
- First posted on arXiv 25 May 2025.

## Placed in

- LLM safety and security / [Vision-language and text-to-image safety](https://mlsanity.com/safety/vision-language-and-text-to-image-safety/)

## Builds on (most cited first, 29 of 29)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Explaining and Harnessing Adversarial Examples](https://mlsanity.com/paper/1412.6572/) (2014)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding](https://mlsanity.com/paper/2205.11487/) (2022)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [Jailbreaking Black Box Large Language Models in Twenty Queries](https://mlsanity.com/paper/2310.08419/) (2023)
- [Universal Adversarial Triggers for Attacking and Analyzing NLP](https://mlsanity.com/paper/1908.07125/) (2019)
- [Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment](https://mlsanity.com/paper/1907.11932/) (2019)
- [Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models](https://mlsanity.com/paper/2211.05105/) (2022)
- [Unified Concept Editing in Diffusion Models](https://mlsanity.com/paper/2308.14761/) (2023)
- [TextBugger: Generating Adversarial Text Against Real-world Applications](https://mlsanity.com/paper/1812.05271/) (2018)
- [BAE: BERT-based Adversarial Examples for Text Classification](https://mlsanity.com/paper/2004.01970/) (2020)
- [Large Scale Crowdsourcing and Characterization of Twitter Abusive Behavior](https://mlsanity.com/paper/1802.00393/) (2018)
- [Red-Teaming the Stable Diffusion Safety Filter](https://mlsanity.com/paper/2210.04610/) (2022)
- [Ring-A-Bell! How Reliable are Concept Removal Methods for Diffusion Models?](https://mlsanity.com/paper/2310.10012/) (2023)
- [MMA-Diffusion: MultiModal Attack on Diffusion Models](https://mlsanity.com/paper/2311.17516/) (2023)
- [SneakyPrompt: Jailbreaking Text-to-image Generative Models](https://mlsanity.com/paper/2305.12082/) (2023)
- [MasterKey: Automated Jailbreak Across Multiple Large Language Model Chatbots](https://mlsanity.com/paper/2307.08715/) (2023)
- [Unsafe Diffusion: On the Generation of Unsafe Images and Hateful Memes From Text-To-Image Models](https://mlsanity.com/paper/2305.13873/) (2023)
- [Can Machines Help Us Answering Question 16 in Datasheets, and In Turn Reflecting on Inappropriate Content?](https://mlsanity.com/paper/2202.06675/) (2022)
- [A Pilot Study of Query-Free Adversarial Attack against Stable Diffusion](https://mlsanity.com/paper/2303.16378/) (2023)
- [SafeGen: Mitigating Sexually Explicit Content Generation in Text-to-Image Models](https://mlsanity.com/paper/2404.06666/) (2024)
- [Black Box Adversarial Prompting for Foundation Models](https://mlsanity.com/paper/2302.04237/) (2023)
- [GuardT2I: Defending Text-to-Image Models from Adversarial Prompts](https://mlsanity.com/paper/2403.01446/) (2024)
- [ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors](https://mlsanity.com/paper/2402.16444/) (2024)
- [Character-level White-Box Adversarial Attacks against Transformers via Attachable Subwords Substitution](https://mlsanity.com/paper/2210.17004/) (2022)
- [Fuzz-Testing Meets LLM-Based Agents: An Automated and Efficient Framework for Jailbreaking Text-To-Image Generation Models](https://mlsanity.com/paper/2408.00523/) (2024)

## Built on by (most cited first, 5 of 5)

- [JANUS: A Lightweight Framework for Jailbreaking Text-to-Image Models via Distribution Optimization](https://mlsanity.com/paper/2603.21208/) (2026)
- [A Systematic Review of Algorithmic Red Teaming Methodologies for Assurance and Security of AI Applications](https://mlsanity.com/paper/2602.21267/) (2026)
- [ReLAPSe: Reinforcement-Learning-trained Adversarial Prompt Search for Erased concepts in unlearned diffusion models](https://mlsanity.com/paper/2602.00350/) (2026)
- [Whispers in the Noise: Surrogate-Guided Concept Awakening via a Multi-Agent Framework](https://mlsanity.com/paper/2605.18150/) (2026)
- [Jailbreaking Text-to-Image Models Through Cracks: Navigating Heterogeneous Safety Filters via Multi-Agent Debate](https://mlsanity.com/paper/2609.01168/) (2026)

## Abstract (arXiv)

Text-to-image (T2I) models can generate not-safe-for-work (NSFW) content, motivating multi-stage safety pipelines with both text and image filters. Newer LLM-based filters detect latent intent beyond keywords, making token-level perturbation attacks unreliable. Our evaluation further shows that existing jailbreak methods exhibit a sharp trade-off between filter evasion and semantic fidelity, while also requiring excessive queries to succeed. We introduce OptJail, an automated jailbreak framework that combines dynamic prompt optimization with multimodal feedback. It consists of two key components: (i) textitDynamic Optimization, an iterative process that leverages text-filter feedback and semantic consistency to rewrite prompts into adversarial variants; and (ii) textitAdaptive Safety Indicator Injection, which formulates the injection of benign visual cues as a reinforcement learning problem to bypass image-level filters. OptJail achieves state-of-the-art performance, increasing the ShieldLM-7B bypass rate from 8.9% (Sneakyprompt) to 99.0%, improving CLIP score from 0.2637 to 0.2762. Moreover, it generalizes to unseen filters and successfully jailbreaks DALL E 3 in our evaluation. Mechanistic analysis reveals why these defenses fail: optimized prompts are projected into the ``safe'' region of the filter's representation space yet remain nearly stationary in the generative model's semantic space, and injected safety indicators redirect image detectors' attention away from NSFW content toward benign visual cues. This study reveals systemic vulnerabilities in current multimodal defenses and motivates stronger adaptive defenses.