Backend
The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer
arXiv cs.SD0 views
arXiv:2602.02557v3 Announce Type: replace-cross Abstract: 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.
Source:https://arxiv.org/abs/2602.02557PreviousZONOS3-Mini Technical ReportNextPerformance-Portable Exact Gaussian Process Regression in Python via SYCL and oneMath
Related posts
The part of the internet AI agents can't access
0dev.to AI
What is an MCP inventory?
0dev.to AI
My Agent Picks a Free AI Model for You. How Do You Know It's Telling the Truth?
0dev.to AI

Zapier vs Make vs n8n (2026): What the Same Workflow Really Costs
0dev.to AI
Gemini 4 Hydrogen, Vercel Gateway, and Bedrock Web Search: Signal #135
0dev.to AI

9 Best Agent Gateways
0dev.to AI