[论文解读] Regulating ChatGPT and other Large Generative AI Models

Philipp Hacker, Andreas Engel+1 more|arXiv (Cornell University)|Feb 5, 2023

Artificial Intelligence in Healthcare and Education被引用 68

一句话总结

本文分析如何监管像 ChatGPT 这样的大型生成式 AI 模型，提出一个框架，对 AI 价值链中不同参与方量身定制职责，并强调透明性、风险管理与内容审核。

ABSTRACT

Large generative AI models (LGAIMs), such as ChatGPT, GPT-4 or Stable Diffusion, are rapidly transforming the way we communicate, illustrate, and create. However, AI regulation, in the EU and beyond, has primarily focused on conventional AI models, not LGAIMs. This paper will situate these new generative models in the current debate on trustworthy AI regulation, and ask how the law can be tailored to their capabilities. After laying technical foundations, the legal part of the paper proceeds in four steps, covering (1) direct regulation, (2) data protection, (3) content moderation, and (4) policy proposals. It suggests a novel terminology to capture the AI value chain in LGAIM settings by differentiating between LGAIM developers, deployers, professional and non-professional users, as well as recipients of LGAIM output. We tailor regulatory duties to these different actors along the value chain and suggest strategies to ensure that LGAIMs are trustworthy and deployed for the benefit of society at large. Rules in the AI Act and other direct regulation must match the specificities of pre-trained models. The paper argues for three layers of obligations concerning LGAIMs (minimum standards for all LGAIMs; high-risk obligations for high-risk use cases; collaborations along the AI value chain). In general, regulation should focus on concrete high-risk applications, and not the pre-trained model itself, and should include (i) obligations regarding transparency and (ii) risk management. Non-discrimination provisions (iii) may, however, apply to LGAIM developers. Lastly, (iv) the core of the DSA content moderation rules should be expanded to cover LGAIMs. This includes notice and action mechanisms, and trusted flaggers. In all areas, regulators and lawmakers need to act fast to keep track with the dynamics of ChatGPT et al.

研究动机与目标

• 促使对大型生成式 AI 模型（LGAIMs）的监管超越传统 AI 监管。
• 区分 AI 价值链中的参与方（开发者、部署者、用户、受众）并据此定制职责。
• 提出多层监管方法（最低标准、高风险义务、价值链各方的协作）。
• 使监管提案与既有框架（AI Act、DSA）保持一致，同时回应 LGAIMs 的独特能力。

提出的方法

• 阐明 LGAIMs 的技术基础，为监管讨论奠定基础。
• 为 LGAIMs 开发新术语以及价值链区分（开发者、部署者、专业/非专业用户、受益者）。
• 按参与方类型和用例在价值链上定制义务。
• 主张三层义务（最低标准、高风险、协作），并将重点放在高风险应用上，而非预训练模型。
• 提出透明度、风险管理、非歧视条款，以及扩展的类似 DSA 的内容审核，配有通知机制与可信标注方。

实验结果

研究问题

• RQ1 应如何对 LGAIMs 的直接监管、数据保护、内容审核与政策进行校准？
• RQ2 应对 LGAIM 价值链中不同参与方施以哪些监管义务？
• RQ3 如何对现有制度（AI Act、DSA）进行调整以应对 LGAIMs 的特殊性？
• RQ4 对 LGAIM 部署与使用，哪些具体措施（透明度、风险管理、非歧视）是合适的？
• RQ5 通知机制和可信标注方在 LGAIM 内容审核中的作用是什么？

主要发现

• 建议将监管聚焦于具体的高风险应用，而不是直接对预训练模型本身进行监管。
• 倡导三层义务结构：最低标准、高风险义务，以及贯穿价值链的跨领域协作。
• 建议将透明度与风险管理纳入 LGAIMs 的核心监管要求。
• 允许将非歧视条款适用于 LGAIM 开发者。
• 将核心 DSA 内容审核规则扩展以覆盖 LGAIMs，具备通知、行动和可信标注方等机制。