Assessing the impact of artificial intelligence systems on fundamental rights

This paper presents a structured framework for conducting a Fundamental Rights Impact Assessment (FRIA) tailored to high-risk Artificial Intelligence (AI) systems, including Generative AI. Grounded in the obligations of the Artificial Intelligence Act, the proposed framework - FRIAct - operationalizes fundamental rights protection by bridging regulatory requirements with practical implementation. The paper situates the FRIA within the broader European legal and constitutional framework, emphasizing the role of the Charter of Fundamental Rights of the European Union in shaping AI governance. It then introduces FRIAct, a two-pronged assessment methodology combining qualitative and quantitative tools, namely a Questionnaire and a Risk Matrix. The effectiveness of the framework in assessing transparency, fairness, privacy, and human oversight is then demonstrated by applying it to a fictitious but realistic case study that represents a high-risk AI system for credit scoring.

Palavras-chave

Fundamental rights Artificial intelligence Artificial intelligence act Risk assessment AI ethics Responsible AI

URI

http://hdl.handle.net/10400.14/59005

Editora

SSRN

DOI

10.2139/ssrn.5168579

Coleções

CEID - Working Papers / Preprints

Licença CC

Sem licença CC

Métricas Alternativas

Ver registo completo