从失控到安全的 AI：显式拒绝在使 LLM 与国际人道法保持一致中的作用
From Rogue to Safe AI: The Role of Explicit Refusals in Aligning LLMs with International Humanitarian Law
AI总结：
本文评估八个主流 LLM 对违反国际人道法请求的拒绝与解释能力，发现标准化系统安全提示可提升多数模型的拒绝解释质量，但复杂技术或代码请求仍存在漏洞。
摘要：
大语言模型（LLM）已在各个领域得到广泛应用，但它们与国际人道法（IHL）的一致性仍未得到充分理解。本研究评估了八个领先 LLM 拒绝明确违反这些法律框架的提示的能力，同时关注有用性，即拒绝被传达得多么清晰和具有建设性。尽管大多数模型拒绝了非法请求，但其回答的清晰度和一致性存在差异。通过揭示模型的基本理由并引用相关法律或安全原则，解释性拒绝能够阐明系统边界、减少歧义，并帮助防止误用。一个标准化的系统级安全提示显著提高了大多数模型在拒绝中所表达解释的质量，凸显了轻量级干预措施的有效性。然而，涉及技术语言或代码请求的更复杂提示暴露出持续存在的漏洞。这些发现有助于开发更安全、更透明的 AI 系统，并提出了一个用于评估 LLM 对 IHL 合规性的基准。
英文摘要：
Large Language Models (LLMs) are widely used across sectors, yet their alignment with International Humanitarian Law (IHL) is not well understood. This study evaluates eight leading LLMs on their ability to refuse prompts that explicitly violate these legal frameworks, focusing also on helpfulness - how clearly and constructively refusals are communicated. While most models rejected unlawful requests, the clarity and consistency of their responses varied. By revealing the model's rationale and referencing relevant legal or safety principles, explanatory refusals clarify the system's boundaries, reduce ambiguity, and help prevent misuse. A standardised system-level safety prompt significantly improved the quality of the explanations expressed within refusals in most models, highlighting the effectiveness of lightweight interventions. However, more complex prompts involving technical language or requests for code revealed ongoing vulnerabilities. These findings contribute to the development of safer, more transparent AI systems and propose a benchmark to evaluate the compliance of LLM with IHL.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
