EU Policy. Commission sets up expert group for AI trade secrets, copyright questions

The AI Act divides AI systems into four main categories according to the potential risk they pose to society.
The AI Act divides AI systems into four main categories according to the potential risk they pose to society. Copyright Markus Schreiber/Copyright 2024 The AP. All rights reserved
Copyright Markus Schreiber/Copyright 2024 The AP. All rights reserved
By Cynthia Kroet
Share this articleComments
Share this articleClose Button

The European Commission has set up an expert group to help the 27 EU member states with questions around the implementation of the AI Act and ensure harmonised application of the rules.

ADVERTISEMENT

The group, consisting of officials drawn from member states’ authorities, should also answer questions anticipated from the countries about product safety, trade secrets, copyright, biometrical identification and law enforcement, Euronews understands from meeting documents. This comes after the AI rulebook was approved at a meeting of EU ambassadors last Friday (2 February).

On product-safety for example, the experts will give advice on the application of the rules in relation to the Medical Device Regulation, In Vitro Diagnostic Device Regulation as well as machinery rules.

Germany expressed support for the working group, the minutes show. Its representative said during the meeting that the link between the AI Act and other legislation needs to be clarified, and had specific questions around law enforcement and border control systems. In turn, France wanted more clarity about trade secrets, whereas Hungary requested more details about law enforcement.

EU countries agreed on the technical details of the AI Act, the world's first attempt to regulate the technology according to a risk-based approach, after a political agreement in December.

The law divides AI systems into four main categories according to the potential risk they pose to society. During the negotiation process, some countries including France, Germany and Italy were sceptical about regulating so-called foundation models such as ChatGPT.

Some of the provisions that leave room for interpretation, or those that require more frequent updates, will need so-called delegated acts, or secondary legislation, to be adopted under the AI Act.

The deal needs a sign-off from EU lawmakers before the rules enter into force; a plenary vote in the European Parliament is expected in April.

After that, the act is expected to enter into force later this year and includes an implementation period of up to 36 months. The requirements for AI models will start to apply already after one year, and within six months EU countries will have to phase out prohibited systems.

In addition to the expert group, the commission is setting up four supervision and enforcement bodies. The AI Office will be tasked with oversight of the implementation, together with national regulators. A European Artificial Intelligence Board, including representatives of each member state, will assist the commission and EU countries and share regulatory experience.

An Advisory Forum will consist of industry representatives, SMEs and academia to give technical input on the AI Act, and finally a Scientific Panel, a pool of independent scientific and technical experts, will help the AI Office in the implementation and enforcement when it comes to General Purpose AI models.

Share this articleComments

You might also like