Making MLLMs Blind: Adversarial Smuggling Attacks in MLLM Content Moderation

• URL: http://arxiv.org/abs/2604.06950v2
• Date: Thu, 09 Apr 2026 01:50:11 GMT
• Title: Making MLLMs Blind: Adversarial Smuggling Attacks in MLLM Content Moderation
• Abstract summary: Adversarial smuggling attacks exploit the Human-AI capability gap.<n>It encodes harmful content into human-readable visual formats that remain AI-unreadable.<n>We conduct a preliminary exploration of mitigation strategies.
• Score: 54.38271718421492
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Multimodal Large Language Models (MLLMs) are increasingly being deployed as automated content moderators. Within this landscape, we uncover a critical threat: Adversarial Smuggling Attacks. Unlike adversarial perturbations (for misclassification) and adversarial jailbreaks (for harmful output generation), adversarial smuggling exploits the Human-AI capability gap. It encodes harmful content into human-readable visual formats that remain AI-unreadable, thereby evading automated detection and enabling the dissemination of harmful content. We classify smuggling attacks into two pathways: (1) Perceptual Blindness, disrupting text recognition; and (2) Reasoning Blockade, inhibiting semantic understanding despite successful text recognition. To evaluate this threat, we constructed SmuggleBench, the first comprehensive benchmark comprising 1,700 adversarial smuggling attack instances. Evaluations on SmuggleBench reveal that both proprietary (e.g., GPT-5) and open-source (e.g., Qwen3-VL) state-of-the-art models are vulnerable to this threat, producing Attack Success Rates (ASR) exceeding 90%. By analyzing the vulnerability through the lenses of perception and reasoning, we identify three root causes: the limited capabilities of vision encoders, the robustness gap in OCR, and the scarcity of domain-specific adversarial examples. We conduct a preliminary exploration of mitigation strategies, investigating the potential of test-time scaling (via CoT) and adversarial training (via SFT) to mitigate this threat. Our code is publicly available at https://github.com/zhihengli-casia/smugglebench.

Related papers

• InceptionRAG: Stealthy Poisoning Attack Against Retrieval-Augmented Generation [23.65420330011369]
  We introduce InceptionRAG, a stealthy attack mechanism that subverts the standard attack paradigm.<n>InceptionRAG achieves an attack success rate exceeding 80% even under rigorous adversarial constraints.<n>Our findings expose a concerning paradox: the stronger reasoning capabilities of LLMs increase their vulnerability to reasoning-based poisoning attacks.
  arXiv  Detail & Related papers  (2026-09-15T08:22:10Z)
• AI Agents May Always Fall for Prompt Injections [13.730410056303866]
  We show that prompt injection is the most critical vulnerability in deployed AI agents.<n>We recast prompt injection via the lens of Contextual Integrity (CI), a privacy theory that judges information flow compliance with contextual norms.
  arXiv  Detail & Related papers  (2026-05-17T19:55:39Z)
• Laundering AI Authority with Adversarial Examples [32.761654180537434]
  Vision-language models (VLMs) are increasingly deployed as trusted authorities.<n>We show that adversarial examples break this assumption, enabling emphAI authority laundering<n>Our attacks do not compromise model alignment; the attack operates entirely at the perceptual level.
  arXiv  Detail & Related papers  (2026-05-05T19:55:28Z)
• Sleeper Cell: Injecting Latent Malice Temporal Backdoors into Tool-Using LLMs [0.0]
  Openweight Large Language Models (LLMs) have democratized agentic AI, yet finetuned weights are frequently shared and adopted with limited scrutiny beyond leaderboard performance.<n>This creates a risk where third-party models are incorporated without strong behavioral guarantees.<n>We show that poisoned models maintain state-of-the-art performance on benign tasks, incentivizing their adoption.
  arXiv  Detail & Related papers  (2026-03-02T22:01:08Z)
• One Token Embedding Is Enough to Deadlock Your Large Reasoning Model [91.48868589442837]
  We present the Deadlock Attack, a resource exhaustion method that hijacks an LRM's generative control flow.<n>Our method achieves a 100% attack success rate across four advanced LRMs.
  arXiv  Detail & Related papers  (2025-10-12T07:42:57Z)
• Revisiting Backdoor Attacks on LLMs: A Stealthy and Practical Poisoning Framework via Harmless Inputs [54.90315421117162]
  We propose a novel poisoning method via completely harmless data.<n>Inspired by the causal reasoning in auto-regressive LLMs, we aim to establish robust associations between triggers and an affirmative response prefix.<n>We observe an interesting resistance phenomenon where the LLM initially appears to agree but subsequently refuses to answer.
  arXiv  Detail & Related papers  (2025-05-23T08:13:59Z)
• Illusions of Relevance: Using Content Injection Attacks to Deceive Retrievers, Rerankers, and LLM Judges [52.96987928118327]
  We find that embedding models for retrieval, rerankers, and large language model (LLM) relevance judges are vulnerable to content injection attacks.<n>We identify two primary threats: (1) inserting unrelated or harmful content within passages that still appear deceptively "relevant", and (2) inserting entire queries or key query terms into passages to boost their perceived relevance.<n>Our study systematically examines the factors that influence an attack's success, such as the placement of injected content and the balance between relevant and non-relevant material.
  arXiv  Detail & Related papers  (2025-01-30T18:02:15Z)
• Turning Logic Against Itself : Probing Model Defenses Through Contrastive Questions [50.40122190627256]
  We introduce POATE, a novel jailbreak technique that harnesses contrastive reasoning to provoke unethical responses.<n>PoATE crafts semantically opposing intents and integrates them with adversarial templates, steering models toward harmful outputs with remarkable subtlety.<n>To counter this, we propose Intent-Aware CoT and Reverse Thinking CoT, which decompose queries to detect malicious intent and reason in reverse to evaluate and reject harmful responses.
  arXiv  Detail & Related papers  (2025-01-03T15:40:03Z)
• Human-Readable Adversarial Prompts: An Investigation into LLM Vulnerabilities Using Situational Context [45.821481786228226]
  We show that situation-driven adversarial full-prompts that leverage situational context are effective but much harder to detect.<n>We developed attacks that use movie scripts as situational contextual frameworks.<n>We enhanced the AdvPrompter framework with p-nucleus sampling to generate diverse human-readable adversarial texts.
  arXiv  Detail & Related papers  (2024-12-20T21:43:52Z)
• You Know What I'm Saying: Jailbreak Attack via Implicit Reference [22.520950422702757]
  This study identifies a previously overlooked vulnerability, which we term Attack via Implicit Reference (AIR) AIR decomposes a malicious objective into permissible objectives and links them through implicit references within the context. Our experiments demonstrate AIR's effectiveness across state-of-the-art LLMs, achieving an attack success rate (ASR) exceeding 90% on most models.
  arXiv  Detail & Related papers  (2024-10-04T18:42:57Z)
• Understanding the Vulnerability of Skeleton-based Human Activity Recognition via Black-box Attack [53.032801921915436]
  Human Activity Recognition (HAR) has been employed in a wide range of applications, e.g. self-driving cars. Recently, the robustness of skeleton-based HAR methods have been questioned due to their vulnerability to adversarial attacks. We show such threats exist, even when the attacker only has access to the input/output of the model. We propose the very first black-box adversarial attack approach in skeleton-based HAR called BASAR.
  arXiv  Detail & Related papers  (2022-11-21T09:51:28Z)
• A Self-supervised Approach for Adversarial Robustness [105.88250594033053]
  Adversarial examples can cause catastrophic mistakes in Deep Neural Network (DNNs) based vision systems. This paper proposes a self-supervised adversarial training mechanism in the input space. It provides significant robustness against the textbfunseen adversarial attacks.
  arXiv  Detail & Related papers  (2020-06-08T20:42:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.