AI "Risks Lie in the Application" Says Germany, France, Italy on AI Regulation Agreement

The nations of France, Italy, and Germany have reached a new artificial intelligence regulation agreement underlining "that the AI Act regulates the application of AI and not the technology as such," as per a report by Reuters.

According to Germany's Economy Ministry, which oversees the matter with the Ministry of Digital Affairs, regulations and governmental authority should not regulate AI itself but its implementation.

Volker Wissing, the Digital Affairs Minister,, states that the agreement was to limit only the use of artificial intelligence. Wissing added, "we need to regulate the applications and not the technology if we want to play in the top AI league worldwide."

Reuters further reported that Franziska Brantner, State Secretary for Economic Affairs, said it was critical to capitalize on possibilities while minimizing risks."

We have developed a proposal that can ensure a balance between both objectives in a technological and legal terrain that has not yet been defined," Brantner said.

AI Regulation Through 'Model Cards'

The joint agreement document states that model cards, which offer information about a machine learning model, would have to be defined by foundation model developers.

"An AI governance body could help to develop guidelines and could check the application of model cards," the three nations stated. "The model cards shall include the relevant information to understand the functioning of the model, its capabilities and its limits and will be based on best practices within the developers community," according to the report.

The three countries support "mandatory self-regulation through codes of conduct" for AI foundation models, which are intended to deliver a wide range of outcomes. They are, however, opposed to "untested norms."

According to the document, no punishments should be enforced at first. However, if violations of the code of conduct are discovered after a specific time, a penalty system might be established.

All-encompassing AI Regulation

This reinforces that the three countries support optional binding obligations for both significant and small AI providers in the European Union, as per a CNBC report on the same matter.

During the debates, the European Parliament recommended that the code of conduct initially apply exclusively to significant AI providers, notably from the United States.

However, the three EU countries have warned against this apparent competitive advantage for smaller European suppliers. According to them, this might lead to reduced faith in the security of these smaller suppliers and hence fewer consumers.

The countries went on to say that the norms of behavior and openness should be binding on everyone.

Reuters reports that the German government will organize a digital summit in Jena, Thuringia on Monday and Tuesday, bringing together officials from politics, industry, and research.

AI issues will also be discussed when the German and Italian governments meet in Berlin on Wednesday.