EU AI ACT: Shaping the Future of Artificial Intelligence in Europe

By Francesca Busato, Federica Antoniazzi

The Regulation (EU) 2024/1689 on Artificial Intelligence marks a pivotal moment in the global regulation of AI, introducing a detailed and harmonized regulatory framework to ensure the responsible and safe use of this technology, which impact on our lives has significantly increased in recent years. The regulation

establishes clear guidelines for AI developers and operators, addressing the risks associated with AI applications and defining specific obligations for those involved with high-risk technologies, such as those used in critical infrastructure, healthcare, education, and justice. A central aspect of the regulation is the risk-based approach, which categorizes AI systems into four main categories:

Unacceptable Risk: Systems deemed a threat to safety or fundamental rights and are therefore prohibited. For example, the use of AI for real-time biometric mass surveillance is strictly forbidden. An example of an AI system posing an unacceptable risk in the legal field is one used to aid judicial decision-making.

High Risk: Systems that present significant risks to safety or rights but may be used if they comply with strict legal requirements, such as certifications, audits, and transparency. This includes, for instance, the use of AI in personnel recruitment or healthcare systems. People will have the right to bring complaints against these kinds of systems before the designated national authorities.

Limited Risk: Systems that require minimal transparency measures, such as notifying users that they are interacting with AI (e.g., chatbots).

Minimal or No Risk: Low-risk AI systems, such as spam filters or personalized suggestions, for which no specific restrictions are provided.

In order to ensure safety and the protection of fundamental rights, the Regulation aims to support innovation and investment in AI while reducing administrative and financial burdens on businesses, particularly SMEs. The regulation is part of a broader package of measures, including the coordinated AI

plan and the AI innovation package, aimed at strengthening the adoption and development of reliable AI across the EU.

Despite its good intentions, the EU Regulation 2024/1689 has raised some concerns. Critics argue that the regulatory framework might slow down innovation, especially in this highly dynamic sector. Others fear that

the regulation could create a gap between Europe and other technological powers, such as the United States and China, where AI regulations are less stringent.

On the other hand, supporters of the regulation assert that strong regulation is necessary to prevent abuses and ensure that AI development aligns with European values of human rights protection and social welfare.

The Regulation (EU) 2024/1689 came into force on August 1, 2024, and will be fully applicable within two years. Some provisions, such as bans on specific AI applications and governance rules, will take effect after six and twelve months, respectively. The EU has also established the European Office for AI, which task will be to oversee the implementation of the law and promoting international cooperation on AI governance.

A further legislative innovation is introduced in Article 112 of the EU AI Act, which allocates resources in order to provide a continuous review and adaptation to the risk level classifications and establishes requirements to keep pace with emerging technologies and their resulting needs. For instance, the most

advanced models, like GPT-4, will need to undergo periodic evaluations, and every incident must be reported to the European Commission.

The EU Regulation 2024/1689 represents a significant step towards creating a secure and reliable digital ecosystem within the European Union. While challenges will undoubtedly arise, its implementation couldset a global standard for AI regulation, balancing innovation with the protection of fundamental rights. The real challenge will be ensuring that this new legislation is effectively enforced, promoting economic growth without compromising the EU's essential values.