France, Germany, and Italy Propose Binding Rules for AI Providers in the EU
France, Germany, and Italy have jointly proposed regulations to govern the use of artificial intelligence (AI) within the European Union (EU). In a recently released joint paper, the three governments advocate for a system of mandatory self-regulation, rather than imposing binding rules on foundational AI models. The proposed regulations would be voluntary but binding for AI providers in the EU that choose to sign up to them. However, the initial proposal does not include any specific sanctions for non-compliance.
The European Parliament had previously introduced the AI Act in June with the aim of managing the risks associated with AI applications and preventing any discriminatory effects, while also promoting innovation. Originally, the code of conduct outlined in the AI Act was intended to be mandatory only for major AI providers, particularly those from the United States. However, France, Germany, and Italy voiced concerns that this approach could undermine trust in smaller European providers and result in fewer customers.
The three countries argue that enforcing horizontal rules for foundational AI models would contradict the technology-neutral and risk-based approach of the AI Act. Instead, they propose that developers of these models should define model cards, which outline the capabilities and limitations of the model. They also suggest that an AI governance body should develop guidelines and review the use of these model cards.
Germany’s Digital Minister, Volker Wissing, emphasized the need to regulate AI applications rather than the technology itself, saying, We need to regulate the applications and not the technology if we want to play in the top AI league worldwide. Additionally, the three governments propose the establishment of a European authority to monitor compliance with the standards. If any violations of the code of conduct are detected after a specified period of time, the authority would make these breaches public to ensure transparency.
The EU is currently in the process of negotiating its stance on AI regulation among the European Commission, the European Parliament, and the EU Council. The joint proposal put forth by France, Germany, and Italy presents an alternative approach to AI regulation within the EU, highlighting the importance of striking a balance between promoting innovation and safeguarding against potential risks.
