The European Commission has taken a significant step toward ensuring the harmonized application of the AI Act across the 27 EU member states by creating an expert group. This group, composed of officials from member states' authorities, will provide guidance on various aspects of the AI Act, including product safety, trade secrets, copyright, biometric identification, and law enforcement.
According to documents from recent meetings, the expert group's responsibilities extend to addressing anticipated questions from member states regarding the application of rules related to product safety, such as those outlined in the Medical Device Regulation, In Vitro Diagnostic Device Regulation, and machinery rules.
Germany has expressed its support for the establishment of this working group, emphasizing the need to clarify the connection between the AI Act and other relevant legislation. During discussions, Germany raised specific inquiries regarding law enforcement and border control systems. Meanwhile, France sought clarification on trade secrets, and Hungary requested additional details on law enforcement matters.
The technical details of the AI Act, which marks the world's first attempt to regulate AI technology using a risk-based approach, have been approved by EU countries following a political agreement in December. The Act categorizes AI systems into four main groups based on the potential risks they pose to society.
To address certain provisions that require interpretation or more frequent updates, the AI Act will be supplemented by delegated acts, or secondary legislation. This agreement awaits approval from EU lawmakers, with a plenary vote in the European Parliament anticipated in April.
Once approved, the AI Act is set to come into force later this year, with an implementation period of up to 36 months. Requirements for AI models will begin to apply within one year, and prohibited systems must be phased out within six months.
In addition to the expert group, the European Commission is establishing four supervision and enforcement bodies. These include the AI Office, tasked with overseeing implementation in collaboration with national regulators, a European Artificial Intelligence Board, an Advisory Forum comprising industry representatives, SMEs, and academia, and a Scientific Panel of independent experts to assist in implementing and enforcing regulations related to General Purpose AI models.
