梳理AI事故治理的痛点，指出当前标准不统一的问题
Open Problems in AI Incident Governance
- 分析现有AI事故治理框架的共性与差异
- 发现定义、分类、报告等环节缺乏一致性
- 适合政策制定者和安全研究人员参考
AI系统在部署后可能产生预部署安全评估无法预见的故障。有效管理这些故障需要健全的AI事故治理体系，包括清晰的定义、分类体系、监测实践、报告机制和事故分析方法。本文考察了监管机构与独立研究者提出的各类框架，发现尽管各框架能描述单一功能的实施方式，但在定义、分类、监测和报告方面仍存在显著不一致。这种不一致体现在事故数据的收集与报告类型、分类方式上，进而影响分析的深度、代表性和准确性。
原文摘要 · Abstract (English)
AI systems may produce failures after deployment that pre-deployment safety assessments do not anticipate. Managing these failures requires what we refer to as adequate \textit{AI incident governance}, where having good definitions, taxonomies, monitoring practices, reporting mechanisms, and incident analysis is essential. We examine existing frameworks related to AI incident governance by regulatory bodies and independent efforts, and find that while there are frameworks that describe how individual functions can be performed, there is a lack of consistency within the aspects of definitions, classification, monitoring, and reporting. These inconsistencies apply to the types of incident data that is collected and reported, the ways in which they are categorised, and as a result, the depth, representativeness, and accuracy of analysis that can be performed.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。