Qualifying algorithmic secrecy

合格的算法保密性

• 批准号: 2751214
• 金额: --
• 依托单位: University of Leeds
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2751214
• 关键词: Qualifying; algorithmic; secrecy

The Role of European Equality Bodies in Balancing Intellectual Property with the Effective Application of Anti-Discrimination LawsThe growing reliance on AI has enabled the identification of a potentially dangerous paradox. While these technologies are capable of reinforcing inequality, the application of equality laws to these discriminatory acts is at risk of being hindered by the presence of algorithmic secrecy enabled by intellectual property. However, despite this acknowledgement, a solution which mitigates such tension without trading-off the interests of either AI developers or citizens is yet to be advanced.Consequently, this project aims to fill in this gap by investigating the extent to which introducing access rights as part of national equality bodies' powers could effectively mitigate the tension between intellectual property and European equality law. Focusing on pre-existing criticism relating to the tension between the proprietary nature of publicly used algorithms and anti-discrimination law, this project will identify transparency as the key factor that limits the effective application of European equality laws. Acknowledging, on the one hand, the growing support for addressing this tension via expertise-based access rights and, on the other, the importance of national equality bodies in preventing and challenging discrimination, this project will seek to assess the merits of integrating such access rights as part of the legal mandate of European equality bodies. This hypothesis, in turn, will be analysed with reference to the necessity of these bodies for such a right, their suitability for this role, and theability of this approach to reflect themes used to justify the strong level of intellectual property protection of algorithms and data.

欧洲平等机构在平衡知识产权与反歧视法的有效适用方面的作用对人工智能的日益依赖使得人们能够识别出一个潜在危险的悖论。虽然这些技术能够加剧不平等，但对这些歧视行为适用平等法律可能会受到知识产权所带来的算法保密的阻碍。然而，尽管如此，在不损害AI开发者和公民利益的前提下缓和这种紧张关系的解决方案还有待推进。因此，本项目旨在通过调查将访问权作为国家平等机构权力的一部分，在多大程度上能够有效缓解知识产权与欧洲平等法之间的紧张关系，从而填补这一空白。该项目聚焦于先前存在的与公共使用算法的专有性质和反歧视法之间的紧张关系有关的批评，将确定透明度是限制欧洲平等法有效应用的关键因素。一方面，认识到越来越多的人支持通过基于专门知识的获取权来解决这一紧张关系，另一方面，认识到国家平等机构在防止和反对歧视方面的重要性，该项目将寻求评估将这种获取权纳入欧洲平等机构的法律的任务的好处。这一假设，反过来，将分析参考这些机构的必要性，这种作用的适合性，以及这种方法的能力，以反映用于证明算法和数据的知识产权保护水平高的主题。