The Hidden Risks of Large Reasoning Models: A Safety Assessment of R1

ACL ARR 2025 February Submission7259 Authors

Abstract:

The rapid development of large reasoning models, such as OpenAI-o3 and DeepSeek-R1, has led to significant improvements in complex reasoning over former non-reasoning large language models~(LLMs). However, their enhanced capabilities, combined with the open-source access of models like DeepSeek-R1, raise serious safety concerns, particularly regarding their potential for misuse.
In this work, we present a comprehensive safety assessment of these reasoning models, leveraging established safety benchmarks to evaluate their compliance with safety regulations across various categories. Furthermore, we investigate their susceptibility to adversarial attacks, such as jailbreaking and prompt injection, to assess their robustness in real-world applications. Our main findings include (1) There is a significant safety gap between the open-source R1 models and the o3-mini model, on both safety benchmark and attack, suggesting more safety effort on R1 is needed. (2) The distilled reasoning model shows poorer safety performance compared to its safety-aligned base models. (3) The stronger the model's reasoning ability, the greater the potential harm it may cause when answering unsafe questions. (4) The reasoning or ``thinking'' outputs in R1 models pose greater safety concerns than their final answers.
Our study provides insights into the security implications of reasoning models and highlights the need for further advancements in R1 models' safety to close the gap.

Paper Type: Long

Research Area: Ethics, Bias, and Fairness

Research Area Keywords: ethical considerations in NLP applications,analysis,large reason models

Contribution Types: Model analysis & interpretability

Languages Studied: English

Submission Number: 7259