Table of Contents
AI regulation: Germany, France and Italy agree on common approach
The future of AI regulation in the EU has become clearer, as Germany, France and Italy have reached a consensus on how to deal with this emerging technology, Reuters reported, citing a joint paper.
The paper, which is expected to speed up the negotiations at the European level, states that the three governments support binding voluntary commitments for all AI providers in the EU, regardless of their size or origin.
EU institutions debate on AI Act to ensure safety and fairness
The EU institutions, namely the European Commission, the European Parliament and the EU Council, are currently discussing how to shape the bloc’s stance on this new field.
The Parliament put forward an “AI Act” in June, which aimed to prevent safety hazards from AI applications and protect against discrimination, while also fostering the innovation potential of this new technology in Europe.
However, the Parliament suggested that the code of conduct should initially apply only to major AI providers, mainly from the U.S.
Code of conduct and transparency for all AI providers in the EU
The three EU governments have opposed this idea, arguing that it would give an unfair advantage to smaller European providers. They said that this could undermine the confidence in the security of these providers and result in fewer customers.
Therefore, they proposed that the rules of conduct and transparency should be mandatory for everyone.
According to the paper, no sanctions should be imposed at first.
But, if breaches of the code of conduct are detected after a certain period of time, a sanction system could be established. A European authority would oversee the adherence to the standards, the paper said.
State should not regulate AI itself, but its use, says German ministry
The German Economy Ministry, which is responsible for the issue along with the Ministry of Digital Affairs, said that the state should not regulate AI itself, but rather its use.
The state should not interfere with the development of AI models that are not yet in use, or have not yet entered the market, the ministry said.
German digital summit and bilateral talks to address AI issues
The German government is hosting a digital summit in Jena, in the state of Thuringia, on Monday and Tuesday, which will bring together representatives from politics, business and science.
AI-related issues will be on the agenda when the German and Italian governments meet in Berlin on Wednesday.
Who will monitor compliance with the standards?
According to a joint paper released by Germany, France, and Italy, a European authority would monitor compliance with AI regulation standards in the EU. This authority would be part of the paper’s proposed governance structure at the European and national levels.
The European Commission has also proposed the creation of a European Artificial Intelligence Board to oversee the implementation of the AI regulation and ensure consistency across the EU. This body would be in charge of issuing opinions and recommendations on issues of AI regulation as they arise, as well as advising national authorities.
What is the European Artificial Intelligence Board?
The European Artificial Intelligence Board (EAIB) is a proposed body that would oversee the implementation of the EU Artificial Intelligence Act, a draught AI regulation aimed at establishing harmonized rules for AI systems in the EU.
The EAIB would be made up of representatives from each EU member state, as well as the European Data Protection Supervisor and the European Commission. It would advise and assist the Commission, as well as contribute to effective cooperation among national supervisory authorities.
According to Article 58 of the proposed AI regulation, the EAIB’s responsibilities would include gathering and sharing expertise and best practises among Member States, as well as contributing to uniform interpretation and administrative practises. The EAIB would also provide the Commission with opinions, recommendations, or written contributions on AI-related issues such as technical specifications, standards, or guidance documents.
