Position: Building Guardrails for Large Language Models Requires Systematic Design

Dong, Y ORCID: 0000-0003-3047-7777, Mu, R, Jin, G, Qi, Y, Hu, J, Zhao, X ORCID: 0000-0002-3474-349X, Meng, J, Ruan, W and Huang, X ORCID: 0000-0001-6267-0366 (2024) Position: Building Guardrails for Large Language Models Requires Systematic Design .

Text 5402_Position_Building_Guardra.pdf - Author Accepted Manuscript Available under License Creative Commons Attribution. Download (8MB) | Preview

Abstract

As Large Language Models (LLMs) become more integrated into our daily lives, it is crucial to identify and mitigate their risks, especially when the risks can have profound impacts on human users and societies. Guardrails, which filter the inputs or outputs of LLMs, have emerged as a core safeguarding technology. This position paper takes a deep look at current open-source solutions (Llama Guard, Nvidia NeMo, Guardrails AI), and discusses the challenges and the road towards building more complete solutions. Drawing on robust evidence from previous research, we advocate for a systematic approach to construct guardrails for LLMs, based on comprehensive consideration of diverse contexts across various LLMs applications. We propose employing sociotechnical methods through collaboration with a multi-disciplinary team to pinpoint precise technical requirements, exploring advanced neural-symbolic implementations to embrace the complexity of the requirements, and developing verification and testing to ensure the utmost quality of the final product.

Item Type:	Conference Item (Unspecified)
Divisions:	Faculty of Science & Engineering Faculty of Science & Engineering > School of Computer Science & Informatics Faculty of Science & Engineering > School of Computer Science & Informatics > Artificial Intelligence Faculty of Science & Engineering > School of Computer Science & Informatics > School of Computer Science & Informatics
Depositing User:	Symplectic Admin
Date Deposited:	05 Dec 2025 09:43
Last Modified:	05 Dec 2025 09:43
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3195873
Disclaimer:	The University of Liverpool is not responsible for content contained on other websites from links within repository metadata. Please contact us if you notice anything that appears incorrect or inappropriate.