The increasing availability of administrative data and recent advances in machine learning, coupled with political support for artificial intelligence, are paving the way for algorithmic decision-making systems in EU administrative procedures. Despite previous efforts to address mutual recognition challenges, businesses still struggle to export products across EU Member States. The advent of the Fourth Industrial Revolution and the potential of artificial intelligence prompt us to consider whether AI systems could help overcome enforcement deficits in certain policy areas.
Using the deployment of AI systems for mutual recognition enforcement as a case study, we explore whether current EU legal frameworks support the automated enforcement of EU law in the internal market. Questions regarding the extent of EU competence in implementing AI systems and the methods of exercising such competence require clarification.
The analysis involves evaluating AI systems deployment from two angles: first, in terms of EU constitutional principles, including competence and transparency in lawmaking; second, at the administrative level, considering safeguards derived from good administration principles and effective judicial protection outlined in the European Charter of Fundamental Rights, as well as general EU legal principles. Additionally, the Mutual Recognition Regulation sets standardised procedural rules and imposes specific obligations on national authorities, emphasising the duty to provide reasons. AI system legality within EU law will be assessed against these established rights and principles.