Respect for Human Rights in European Legislation on the Use of Artificial Intelligence

European regulation of artificial intelligence seeks to reconcile two objectives that sometimes pull in different directions: on the one hand, the desire to encourage innovation, and on the other, the need to strongly protect fundamental rights (human rights as defined in the European context). At the core of this framework is Regulation (EU) 2024/1689, which organizes obligations based on risk: it prohibits certain practices deemed unacceptable, imposes pre- and postdeployment requirements for high-risk systems, and requires transparency for uses where the risk is lower. The AI Act is grounded in the values and rights of the EU Charter of Fundamental Rights and operates alongside existing rules, particularly the GDPR (which it does not replace), including for automated decision-making and impact assessments. The analysis highlights two new elements relevant to rights: (1) the Fundamental Rights Impact Assessment (FRIA), required in certain cases prior to use, particularly in the public sector or in services of public interest; (2) a right to an explanation when an individual decision is based primarily on the output of a high-risk AI system. At the international level, the criteria developed by the European Court of Human Rights—legality, necessity, proportionality, and safeguards against abuse—particularly regarding biometrics and surveillance, serve as a useful benchmark for interpreting and applying the AI Act.