France, Germany and Italy are set to publish a joint paper in which the three European nations agree on how artificial intelligence (AI) should be regulated.

In the paper, seen by Reuters, the three governments agree to support "mandatory self-regulation through codes of conduct" for foundation models of AI while opposing “un-tested norms”.

Specifically, the paper lays out that developers of foundation models must define model cards which are used to provide information about the machine learning model. These cards will lay out the functioning of the model, along with its capabilities and limits.

The paper surmises: "Together we underline that the AI Act regulates the application of AI and not the technology as such. The inherent risks lie in the application of AI systems rather than in the technology itself."

"An AI governance body could help to develop guidelines and could check the application of model cards.”

It also suggests that sanctions could be set up if violations of the code of conduct are identified.

The paper is to be published by the three continental European nations, but its development was led by the German Economy Ministry and the Ministry of Digital Affairs.

Speaking to Reuters, Digital Affairs minister Volker Wissing said: "We need to regulate the applications and not the technology if we want to play in the top AI league worldwide.”

The paper’s publication should go some way to accelerating regulation of AI across Europe, with the European Commission, the European Parliament and the EU Council all currently negotiating how the bloc should handle the issue.

Following the staging of the UK’s first AI safety summit earlier this month at Bletchley Park, the German government is holding a digital summit this week with political and business leaders set to discuss AI as one of the chief topics.

---