Scholarlink

<p>Artificial intelligence offers the potential for faster diagnosis, broader access, and more efficient healthcare delivery across the European Union. However, it also poses risks of reinforcing bias, exclusion, and diminishing patient autonomy. The European Union has responded with a complex array of recent regulatory instruments, including the AI Act (Regulation (EU) 2024/1689), the European Health Data Space Regulation (EU) 2025/327, the General Data Protection Regulation (GDPR), the Medical Device Regulation, the Data Act, and the Data Governance Act. While these instruments are technically advanced, they are primarily structured around system risk, market access, or data flows, rather than focusing on the patient as a rights-holder. In response, this paper proposes a rights-integrated constitutional framework for digital health governance that recenters these instruments on the individual, using the right to health (ICESCR Article 12, as interpreted in CESCR General Comment No. 14) as the central integrating norm. The framework is built on three pillars: accountability, agency, and participation, each anchored in existing legal mechanisms rather than in newly asserted justiciable rights. The principal operational recommendation is to reorient the AI Act’s fundamental-rights impact assessment (Article 27) into a patient-facing, contestable entitlement, granting individuals the right to be informed when an assessment is required, to access the assessment, and to challenge deployments that were not adequately anticipated. The contribution of this approach lies in integrating existing entitlements rather than creating additional ones. The framework is situated within an explicit enforceability taxonomy and addresses key objections, including concerns that the proliferation of rights may undermine enforceability.</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

From compliance to claim: re-orienting the EU’s fundamental rights impact assessment for patient-centred AI in health

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月4日

查看全文
  • Antonio Pele,
  • José Vida Fernández,
  • Francisco Ortega,
  • Caitlin Mulholland,
  • Carolina Bouchardet

摘要

Artificial intelligence offers the potential for faster diagnosis, broader access, and more efficient healthcare delivery across the European Union. However, it also poses risks of reinforcing bias, exclusion, and diminishing patient autonomy. The European Union has responded with a complex array of recent regulatory instruments, including the AI Act (Regulation (EU) 2024/1689), the European Health Data Space Regulation (EU) 2025/327, the General Data Protection Regulation (GDPR), the Medical Device Regulation, the Data Act, and the Data Governance Act. While these instruments are technically advanced, they are primarily structured around system risk, market access, or data flows, rather than focusing on the patient as a rights-holder. In response, this paper proposes a rights-integrated constitutional framework for digital health governance that recenters these instruments on the individual, using the right to health (ICESCR Article 12, as interpreted in CESCR General Comment No. 14) as the central integrating norm. The framework is built on three pillars: accountability, agency, and participation, each anchored in existing legal mechanisms rather than in newly asserted justiciable rights. The principal operational recommendation is to reorient the AI Act’s fundamental-rights impact assessment (Article 27) into a patient-facing, contestable entitlement, granting individuals the right to be informed when an assessment is required, to access the assessment, and to challenge deployments that were not adequately anticipated. The contribution of this approach lies in integrating existing entitlements rather than creating additional ones. The framework is situated within an explicit enforceability taxonomy and addresses key objections, including concerns that the proliferation of rights may undermine enforceability.