[论文解读] Generative Discrimination: What Happens When Generative AI Exhibits Bias, and What Can Be Done About It

Philipp Hacker|arXiv (Cornell University)|Jan 1, 2024

Impact of AI and Big Data on Business and Society被引用 5

一句话总结

本文讨论生成式人工智能如何表现出歧视、对输出进行分类，并提出监管更新与问责机制以减轻偏见。

ABSTRACT

As generative Artificial Intelligence (genAI) technologies proliferate across sectors, they offer significant benefits but also risk exacerbating discrimination. This chapter explores how genAI intersects with non-discrimination laws, identifying shortcomings and suggesting improvements. It highlights two main types of discriminatory outputs: (i) demeaning and abusive content and (ii) subtler biases due to inadequate representation of protected groups, which may not be overtly discriminatory in individual cases but have cumulative discriminatory effects. For example, genAI systems may predominantly depict white men when asked for images of people in important jobs. This chapter examines these issues, categorizing problematic outputs into three legal categories: discriminatory content; harassment; and legally hard cases like unbalanced content, harmful stereotypes or misclassification. It argues for holding genAI providers and deployers liable for discriminatory outputs and highlights the inadequacy of traditional legal frameworks to address genAI-specific issues. The chapter suggests updating EU laws, including the AI Act, to mitigate biases in training and input data, mandating testing and auditing, and evolving legislation to enforce standards for bias mitigation and inclusivity as technology advances.

研究动机与目标

• 研究生成式AI如何与反歧视法相互作用并识别差距。
• 识别两种主要的歧视性输出类型：贬低性内容和表征偏见。
• 提出对生成式AI提供方与部署方的问责机制，并讨论法律框架的不足。

提出的方法

• 将有问题的生成式AI输出分类为歧视性内容、骚扰，以及像内容失衡或错误分类这样的棘手法律案例。
• 讨论生成式AI输出的提供方与部署方的责任影响。
• 审查并提出对欧盟法律的更新建议，包括AI法案，以解决数据与模型输出中的偏见。
• 主张强制测试、审计以及针对偏见缓解和包容性的标准。

实验结果

研究问题

• RQ1 生成式AI可能产生哪些歧视性输出，法律上应如何分类？
• RQ2 针对歧视性输出，生成式AI提供方与部署方的责任应如何划分？
• RQ3 需要哪些法律框架变更以解决生成式AI特有的偏见和包容性？
• RQ4 如何强制实施测试与审计以减轻训练数据和输入数据的偏见？
• RQ5 欧盟法规（如AI法案）在执行偏见缓解标准方面应扮演怎样的角色？

主要发现

• 生成式AI输出可能是歧视性内容、骚扰，或属于诸如内容失衡和有害刻板印象等棘手法律类别。
• 当前反歧视法律与生成式AI特定偏见之间存在不充分的对齐。
• 责任应扩展到生成式AI的提供者和部署者，对歧视性输出承担责任。
• 包括AI法案在内的欧盟法律需要更新，以减轻训练数据和输入数据的偏见，并强制进行测试与审计。
• 随着技术进步，立法应发展以强制执行偏见缓解和包容性标准。