Retrospective of our contribution to the EU AI Act

Observatorio de Riesgos Catastróficos Globales (ORCG), welcomes the adoption of the Artificial Intelligence Regulation the European Union's Artificial Intelligence Regulation. On December 9, the Parliament, the Council, and the Commission announced they had reached a political agreement after months of intense negotiations. The member states of the Council member states member states approved the agreed text on February 2, while the main committees of the European Parliament main committees did the same on February 13. In April, the Parliament will host the final vote in plenary session, after which the regulation will be published in the Official Journal. 

Once officially processed, the European Regulation will be the first law to comprehensively regulate artificial intelligence, becoming a milestone and benchmark in the global governance of this technology. The Observatory congratulates all the actors involved for the commendable task.  We would especially like to congratulate the Spanish Presidency of the Council, whose commitment and mediation have been fundamental for the negotiations to reach a successful conclusion. 

We are pleased to see that the final text incorporates proposals to governing general-purpose models with systemic risk We are pleased to see that the final text incorporates proposals for governing general purpose models with systemic risk, including the evaluation of models, the creation of a risk assessment and mitigation system, or the monitoring of incidents and the adoption of corrective measures to address them. We are also pleased to have contributed with information to the definition of a formula that enables a more precise and balanced regulation, requiring proportional obligations according to specific criteria such as the amount of computation used to train a model. You can consult a selection of our published papers below: 

Figure 1. Timeline of contributions

We consider the Regulation to be an important first step, but its effectiveness will depend on the work that will be done from now on to ensure its implementation and compliance. In this regard, we applaud the decision to establish an European AI OfficeThe European AI Office, a unique agency in the world that will be instrumental in managing the risks posed by the most advanced AI systems. ORCG expresses its willingness to support the role of the Office, as well as that of other actors that will accompany it along the way, such as the Scientific Panel, the AI Board and the Joint Research Center. Cooperation and science will be essential to ensure a safe and beneficial future for all. 

Anterior
Anterior

Artículo académico "Revisión sistemática de taxonomías de riesgos asociados a la Inteligencia Artificial"

Siguiente
Siguiente

Recomendación de políticas: "Resiliencia agrícola en América Latina: enfrentando los Escenarios de Reducción Abrupta de la Luz Solar”