Beyond the Safety Tax: Mitigating Unsafe Text-to-Image Generation via External Safety Rectification

ACL ARR 2026 January Submission9332 Authors

Keywords: Text-to-Image, Safety Alignment, Diffusion Models, Mitigation

Abstract: Text-to-image (T2I) generative models have achieved remarkable visual fidelity, yet remain vulnerable to generating unsafe content. Existing safety defenses typically intervene internally within the generative model, but suffer from severe concept entanglement, leading to degradation of benign generation quality—a trade-off we term the Safety Tax. To overcome this limitation, we advocate a paradigm shift from destructive internal editing to external safety rectification. Following this principle, we propose SafePatch, a structurally isolated safety module that performs external, interpretable rectification without modifying the base model. The core backbone of SafePatch is architecturally instantiated as a trainable clone of the base model’s encoder, allowing it to inherit rich semantic priors and maintain representation consistency. To enable interpretable safety rectification, we construct a strictly aligned counterfactual safety dataset (ACS) for differential supervision training. Across nudity and multi-category bench- marks and recent adversarial prompt attacks, SafePatch achieves robust unsafe suppression (7% unsafe on I2P) while preserving image quality and semantic alignment.

Paper Type: Long

Research Area: Safety and Alignment in LLMs

Research Area Keywords: safety and alignment,text-to-text generation

Contribution Types: Model analysis & interpretability

Languages Studied: English

Submission Number: 9332