Visual Self-Fulfilling Alignment: Shaping Safety-Oriented Personas via Threat-Related Images

Multimodal large language models (MLLMs) face safety misalignment, where visual inputs enable harmful outputs. To address this, existing methods require explicit safety labels or contrastive data; yet, threat-related concepts are concrete and visually depictable, while safety concepts, like helpfulness, are abstract and lack visual referents. Inspired by the Self-Fulfilling mechanism underlying emergent misalignment, we propose Visual Self-Fulfilling Alignment (VSFA). VSFA fine-tunes vision-language models (VLMs) on neutral VQA tasks constructed around threat-related images, without any safety labels. Through repeated exposure to threat-related visual content, models internalize the implicit semantics of vigilance and caution, shaping safety-oriented personas. Experiments across multiple VLMs and safety benchmarks demonstrate that VSFA reduces the attack success rate, improves response quality, and mitigates over-refusal while preserving general capabilities. Our work extends the self-fulfilling mechanism from text to visual modalities, offering a label-free approach to VLMs alignment.

Key Contributions

VSFA: a label-free VLM safety alignment method that fine-tunes on neutral VQA tasks built around threat-related images, requiring no explicit safety labels or contrastive data
Extends the self-fulfilling alignment mechanism from text to the visual modality, enabling safety-oriented persona formation via implicit visual semantics
Empirically demonstrates reduced attack success rate, improved response quality, and mitigated over-refusal across multiple VLMs and safety benchmarks while preserving general capabilities