The Fundamental Rights Impact Assessment (FRIA) under the EU AI Act is meant to protect fundamental rights from AI harms, but its practical rollout raises hard questions. This article unpacks FRIA’s core components and its relationship to the GDPR’s Data Protection Impact Assessment, showing how FRIA’s focus on “interferences” — rather than just quantifiable “damages” — gives it a far broader reach, including cases with no personal data at all. It then turns to the EU’s boldest move: an expansive jurisdictional claim built on “territorializing” extraterritorial obligations and the effects doctrine, positioning FRIA as a tool of internal due diligence with global human rights ambitions. The piece closes with a research agenda addressing sovereignty conflicts, AI’s black-box opacity, and the practical hurdles of enforcing rights-based standards worldwide.

On March 21st, 2024, the United Nations General Assembly adopted for the first time a non-binding Resolution that encourages the development of regulatory and governance principles and frameworks and includes standards to ensure that safe, secure and trustworthy AI systems are created.

On September 7th, 2023, the Court of Justice of the European Union (CJEU) upheld the decision of the General Court according to which the public can partially access documentation on the EU’s emotion recognition project (iBorderCtrl) in which it discusses the general reliability, ethics and legality of such technology.

The purpose of the Convention is to ensure that during their lifecycle, AI systems fully comply with human rights, respect the functioning of democracy and observe the rule of law, regardless of whether these activities are undertaken by public or private actors. The design, development and application of AI systems used for purposes related to national defence are expressly excluded from the scope of this Convention. The negotiators seem to agree that such a Convention must be seen first and foremost as a broad framework which might be supplemented by further obligations in more specific fields.

On December 8th, 2022, the European Union Agency for Fundamental Rights (FRA) published a report on algorithm biases, in particular when used for predictive policing and offensive speech detection.