The media has praised the proposed new EU Artificial Intelligence Act as a bold action against the perceived dangers of emerging computer technology.



It consists of an initial proposal for a Regulation with annexes from 2021, along with recent Amendments adopted on 14 June. The regulatory behemoth is intertwined with other recent EU regulations and consists of nearly 100 "recitals," followed by a hundred clauses and ten annexes. It is a complex and evolving text that continues to change among a set of invisible players.

The AI Act now has over fifty vague definitions. It includes various regulatory mechanisms such as capability requirements, prohibited practices, standards, certifications, monitoring, enforcement, and penalties. The Act's obligations apply to anyone in the world involved in placing AI systems on the market or using them, including providers, importers, distributors, and authorized representatives.

The definition of an AI system is so broad that it encompasses almost any computer-based activity. The AI Act itself is constantly evolving through legislative text and the delegation of details to new activities and organizations.

The purpose of the AI Act is to promote human-centric and trustworthy artificial intelligence while ensuring the protection of health, safety, fundamental rights, democracy, the rule of law, and the environment. However, the current attributes of the Act raise concerns and can be seen as a failure by design.

Flawed legal constructs make the implementation of the AI Act in its present form essentially impossible due to significant legal obstacles.

Jurisdiction and Conflict of Law. Instead of collaborating with other jurisdictions on a common approach, the EU decided to proceed with its own regulatory regime for AI. Nations outside the EU are taking different approaches, notably the ITU, the only global intergovernmental body significantly engaged in AI law and standardization. Last year, nearly all nations added Resolution 214 to the ITU ICT global treaty instrument to address AI technologies, taking a more positive and thoughtful approach. The ITU has brought nations together for various AI activities and developed AI technical standards for radiocommunications and ICT networks, resulting in beneficial AI systems and implementations.