Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonable

Safety alignment is an important procedure before the official deployment of a Large Language Model (LLM). While safety alignment has been extensively studied for LLM, there is still a large research gap for Large Reasoning Models (LRMs) that equip with improved reasoning capability. We in this paper systematically examine a simplified pipeline for producing safety aligned LRMs. With our evaluation of various LRMs, we deliver two main findings: i) Safety alignment can be done upon the LRM to restore its safety capability. ii) Safety alignment leads to a degradation of the reasoning capability of LRMs. The two findings show that there exists a trade-off between reasoning and safety capability with the sequential LRM production pipeline. The discovered trade-off, which we name Safety Tax, should shed light on future endeavors of safety research on LRMs. As a by-product, we curate a dataset called DirectRefusal, which might serve as an alternative dataset for safety alignment. Our source code is available atthis https URL.

@article{huang2025_2503.00555,
  title={ Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonable },
  author={ Tiansheng Huang and Sihao Hu and Fatih Ilhan and Selim Furkan Tekin and Zachary Yahn and Yichang Xu and Ling Liu },
  journal={arXiv preprint arXiv:2503.00555},
  year={ 2025 }
}