New EU Artificial Intelligence Board meets for the first time.

The EU has set itself the goal of tapping the full potential of artificial intelligence (AI), while ensuring its safe and reliable application. This is to be achieved primarily through the Regulation laying down harmonised rules on artificial intelligence (AI Act), which came into force on 1 August 2024. As part of the adoption of the AI Act, a new European Artificial Intelligence Board (AI Board) met officially for the first time on 10 September. Articles 65 and 66 of the AI Act form the basis for the establishment of the AI Board.

Composition of the AI Board

The AI Board comprises representatives of the European Commission and the EU Member States. Moreover, the European Data Protection Supervisor is part of the AI Board as an observer. Representatives of the European AI Office also take part in meetings of the AI Board, but do not participate in voting. Other authorities or bodies of the EU or the Member States may be invited to individual meetings of the AI Board depending on their relevance. The AI Board is chaired by a representative of the Member States.

Organisation of the AI Board

There are two permanent subgroups in the AI Board to provide a platform for cooperation and exchange between national market surveillance authorities and to inform authorities on matters relating to market surveillance and notified bodies. In Germany, the Federal Network Agency has been appointed as the market surveillance authority and the German Accreditation Body as the notifying body. In addition, the AI Board can set up further permanent or temporary subgroups as required.

Tasks of the AI Board

According to the AI Act, the AI Board is to support and advise the Commission and the Member States, so as to facilitate the harmonised and effective application of the law. With this aim in mind, the AI Board can, for example, issue recommendations and written opinions on relevant issues relating to the adoption and application of the AI Act, including the annual review of applications categorised as high-risk AI systems, which are subject to particularly strict regulation under the AI Act.

Impact on social insurance

AI systems that are used in the areas of accessibility and recourse to essential private and public services and benefits are classified as high-risk AI systems. This applies, for example, to AI systems that regulate access to healthcare or rehabilitation services. The regulations for such systems include the establishment of a risk management system as well as quality management and information obligations. Applications used by individual social insurance agencies do not yet fall into this category of high-risk AI systems. Nevertheless, this could change in the future if machine learning procedures are used to review a benefit claim, for instance.