对齐诅咒：模态对齐通过文本传输增强音频攻击
The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer
AI总结：
本文提出并验证了“对齐诅咒”原理，即更强的文本-音频模态对齐会促进文本攻击向音频的迁移，并通过黑盒实验表明文本转移的音频攻击性能与原生音频攻击相当甚至更优，揭示了能力与安全之间的根本矛盾。
摘要：
近期端到端训练的全能模型通过加强文本-音频模态对齐显著提升了音频能力。然而，这种对齐是否无意中促进了安全漏洞跨模态的转移仍未被充分探索。这一问题至关重要，因为基于文本的越狱攻击远比基于音频的攻击成熟；如果它们系统性转移，当前的音频安全评估可能低估源自文本模态的风险。在本文中，我们引入了“对齐诅咒”，这是一个经过形式化表征和实证验证的原理，表明更强的模态对齐使得攻击从文本到音频的转移更有效，揭示了能力与安全之间的根本矛盾。基于这一原理，我们在最新的全能模型（如Qwen2.5-Omni、Qwen3-Omni）上对三类攻击（文本攻击、文本转移的音频攻击和音频攻击）进行了全面的黑盒评估。我们发现，文本转移的音频攻击与基于音频的攻击表现相当，甚至更优，在仅音频访问下展现出明显优势。这表明基于文本的漏洞在塑造音频安全风险中扮演关键角色。最后，我们实证分析了不同攻击方法和模型下模态对齐与转移有效性之间的关系，观察到对“对齐诅咒”的一致支持：更紧密的模态对齐导致更有效的跨模态攻击转移。
英文摘要：
Recent advances in end-to-end trained omni-models have substantially improved audio capabilities by strengthening text-audio modality alignment. However, whether such alignment inadvertently facilitates the transfer of safety vulnerabilities across modalities remains underexplored. This question is critical as text-based jailbreak attacks are considerably more mature than audio-based ones; if they transfer systematically, current audio safety evaluations may underestimate risks originating from the text modality. In this paper, we introduce the Alignment Curse, a formally characterized and empirically validated principle showing that stronger modality alignment enables more effective transfer of attacks from text to audio, revealing a fundamental tension between capability and safety. Motivated by this principle, we conduct a comprehensive black-box evaluation of three attack categories on recent omni-models (e.g., Qwen2.5-Omni, Qwen3-Omni): text attacks, text-transferred audio attacks, and audio attacks. We find that text-transferred audio attacks perform comparably to, and often better than, audio-based attacks, exhibiting a clear advantage under audio-only access. This suggests that text-based vulnerabilities play a pivotal role in shaping audio safety risks. Finally, we empirically analyze the relationship between modality alignment and transfer effectiveness across attack methods and models, observing consistent support for the Alignment Curse: tighter modality alignment leads to more effective cross-modality attack transfer.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
