The inclusion and participation of actors involved in artificial intelligence governance applied to public administrative systems and procedures

The primary objective was to build a model that complements the provisions of the recent European Union AI regulation, the National Institute of Standards and Technology (NIST) Artificial Intelligence Risk Management Framework—which operationalizes the US President’s Executive Order 14110—and the first international standard for the Artificial Intelligence Management Systems (ISO/IEC 42001:2023) of the International Standardization Organization (ISO). This objective was a priority because Responsibility Articles 28 and 57 of the recent European Union regulations have set a deadline of August 2025 and August 2026 for designating an authority responsible for evaluations and testing before artificial intelligence (AI) systems are put into service. The previous objective analyzes the above regulations and provisions from the perspective of AI governance (AIG). That is, the approach seeks to balance the empowerment of algorithms in public administration with the citizen aspiration of empowerment. The method analytically reviews the stages and actions of an AI system to infer the design and development of the action and responsibilities of the actors involved in the AIG process from end to end. The results show a general AIG model for public administration that uses Artificial Intelligence before, during, and after its complete life cycle. The conclusions demand a holistic vision that includes both social and technical infrastructure.