{"ID":3004744,"CreatedAt":"2026-06-03T03:09:48.883664427Z","UpdatedAt":"2026-06-05T11:43:53.432517148Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2606.03793","arxiv_id":"2606.03793","title":"Exploring Adversarial Robustness and Safety Alignment in Multilingual Multi-Modal Large Language Models","abstract":"Multimodal Large Language Models integrate visual perception into language reasoning, introducing a continuous attack surface susceptible to adversarial attacks. Prior work on MLLM robustness has focused largely on English-centric tasks, leaving multilingual behaviour unexplored. We address this gap through a systematic study of adversarial robustness and multimodal safety across 12 diverse languages, evaluating open-source MLLMs that acquire multilingual capability through instruction tuning. Gradient-based attacks reveal a transferable multilingual vulnerability: adversarial images optimized in one language continue to induce failure in others, demonstrating strong cross-lingual transferability. Multilingual safety further varies with how effectively a model retrieves or interprets harmful instructions. When harmful intent is issued through text, languages with stronger linguistic grounding more often elicit misuse-enabling responses, while weaker languages produce fewer unsafe outputs. When embedded in the image as typographic content, English scripts are reliably recognised and followed, whereas non-English scripts are rarely parsed by the vision encoder. Lower-resource languages may therefore appear safer, but this is an artefact of comprehension and visual-grounding failures rather than genuine alignment, a phenomenon we term safety-by-failure. In contrast, MLLMs that build multilingual capability throughout their training stages rather than only at instruction tuning, such as Qwen3-VL, exhibit genuine cross-lingual safety, maintaining active refusal across languages rather than masking comprehension failure. Shallow multilingual adaptation, such as fine-tuning on translated instruction data, may produce surface-level understanding that creates illusory safety in low-resource languages; deeper integration across training stages leads to genuine multilingual safety alignment.","short_abstract":"Multimodal Large Language Models integrate visual perception into language reasoning, introducing a continuous attack surface susceptible to adversarial attacks. Prior work on MLLM robustness has focused largely on English-centric tasks, leaving multilingual behaviour unexplored. We address this gap through a systemati...","url_abs":"https://arxiv.org/abs/2606.03793","url_pdf":"https://arxiv.org/pdf/2606.03793v1","authors":"[\"Hashmat Shadab Malik\",\"Muzammal Naseer\",\"Salman Khan\"]","published":"2026-06-02T15:42:10Z","proceeding":"cs.CL","tasks":"[\"cs.CL\",\"cs.CV\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}