Scholarlink

<p>We propose autonomy, solidarity, and justice as a practical consolidation of the EU High-Level Expert Group’s trustworthy AI framework for evaluating the implementation of the AI Act. This triadic lens simplifies complex ethical requirements while maintaining operational clarity. Our two contributions are as follows: first, a framework that maps these three values onto existing regulatory instruments (GDPR, DSA, AI Act), showing how each generates distinct compliance pathways; second, concrete implementation mechanisms through AI Ethics Foresight Councils (AEFCs) and Value-Rights Indicators (VRIs), which transform abstract principles into measurable outcomes. We demonstrate application through analysis of the AI Act’s law enforcement provisions, where categorical prohibitions operationalise justice, transparency requirements enable autonomy, and collective safeguards embody solidarity. AEFCs would provide anticipatory guidance through quarterly horizon scanning, citizen panels, and non-binding opinions linked to enforcement via explain-or-comply mechanisms. VRIs translate values into auditable metrics integrated within Fundamental Rights Impact Assessments and conformity procedures. This framework offers policymakers practical tools for navigating AI governance complexity while safeguarding fundamental rights.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Operationalising ethical AI governance: a triadic framework for the European AI act

  • 期刊论文

  • 发表日期：

    2025年12月1日

查看全文
  • Joffrey Baeyaert

摘要

We propose autonomy, solidarity, and justice as a practical consolidation of the EU High-Level Expert Group’s trustworthy AI framework for evaluating the implementation of the AI Act. This triadic lens simplifies complex ethical requirements while maintaining operational clarity. Our two contributions are as follows: first, a framework that maps these three values onto existing regulatory instruments (GDPR, DSA, AI Act), showing how each generates distinct compliance pathways; second, concrete implementation mechanisms through AI Ethics Foresight Councils (AEFCs) and Value-Rights Indicators (VRIs), which transform abstract principles into measurable outcomes. We demonstrate application through analysis of the AI Act’s law enforcement provisions, where categorical prohibitions operationalise justice, transparency requirements enable autonomy, and collective safeguards embody solidarity. AEFCs would provide anticipatory guidance through quarterly horizon scanning, citizen panels, and non-binding opinions linked to enforcement via explain-or-comply mechanisms. VRIs translate values into auditable metrics integrated within Fundamental Rights Impact Assessments and conformity procedures. This framework offers policymakers practical tools for navigating AI governance complexity while safeguarding fundamental rights.