top of page

AI Act: Towards an Innovative and Secure Future in Artificial Intelligence

The European Union is preparing to launch cutting-edge legislation in the field of Artificial Intelligence, which promises to be the most complete and advanced on a global level.

The primary objective of this legislation is to promote the development of artificial intelligence while ensuring full respect for fundamental human rights . After exhausting negotiations, which lasted for over 36 hours over three days, the three main European institutions - the European Parliament, the European Commission and the European Council - reached a political agreement on the so-called AI Act. Although the text requires further refinements in the coming weeks, the agreement reached guarantees that the legislation will be ratified by the end of the current parliamentary mandate, and then gradually implemented over the next two years. This result was not at all obvious, especially considering the divergent initial positions: on the one hand, the European Parliament, more inclined towards the protection of civil rights, and on the other, the national governments, more oriented towards the needs of economic development and order public.

One of the most delicate points of the negotiations concerned the regulation of the most advanced artificial intelligence, such as those developed by leading companies such as OpenAi, Meta and Google. On Thursday, the obstacle relating to these technologies was overcome: they will be subjected to binding regulations on transparency and security, rather than simple codes of conduct on a voluntary basis. Furthermore, a compromise was reached on the last critical point, namely the use of artificial intelligence applications in the police sector: biometric recognition, to mention one of the most controversial aspects, will only be permitted in cases of serious crimes and subject to judicial authorization .

Thierry Breton, European Commissioner, celebrated this moment by calling it historic and emphasized how the AI Act represents a springboard for researchers and companies Europeans, allowing them to position themselves at the forefront of the global AI race. Brando Benifei, MEP of the Democratic Party and rapporteur of the legislation, underlined how this law places rights and freedoms at the center of the development of this revolutionary technology, balancing innovation and protection. However, numerous uncertainties remain regarding the final balance, application and effectiveness of the AI Act.

The core of the AI Act lies in the classification of artificial intelligence applications based on the level of risk they represent to fundamental rights. In this context, some areas considered excessively risky are prohibited, such as social scoring systems, already theorized and tested in some countries, and those that manipulate human decisions and behaviors. Emotion recognition algorithms will be banned from schools and workplaces, while they appear to be able to find application in contexts related to immigration and security, as requested by governments. The European Parliament achieved a significant victory regarding categorization systems based on sensitive data, such as race, religion, sexual orientation, which will be banned.

The AI Act, the legislative proposal of the European Union, presents itself as an ambitious step forward in addressing the challenges posed by intelligence artificial. This law aims to be a global model for the regulation of AI, putting human rights and security at the center, without neglecting innovation and economic growth.

The regulation is divided into various levels, dividing AI applications into categories based on the risk they present. Low-risk applications, such as chatbots and recommendation algorithms, will be subject to minimal transparency requirements. For intermediate risk technologies, such as those used in the healthcare or transportation sectors, greater guarantees in terms of safety and reliability will be required.

A crucial aspect of the legislation concerns the processing of personal data. The AI Act introduces strict restrictions on the use of biometric data, such as fingerprints or facial recognition, limiting their use to specific cases and under strict control. Furthermore, it requires that AI algorithms used for profiling and for decisions that have a significant impact on people's lives be transparent and explainable, so as to guarantee a right of appeal.

Measures to promote innovation are also planned. The AI Act encourages the creation of "sandboxes" regulatory, spaces where companies can experiment with AI solutions in a controlled environment, receiving support and guidance from the authorities. This approach aims to balance the need for regulation with that of not stifling technological progress.

The AI Act also provides for a supervision and control mechanism, with the creation of a European body which will have the task of monitoring the implementation of the law and intervening in case of violations. Penalties for companies that do not comply with the regulations will be severe, with fines that can reach 4% of global turnover.

Reaction to the new regulation has been mixed. Some experts and digital rights activists have praised the EU's holistic approach, while others have criticized it for not being tough enough on big tech companies. On the other hand, the tech industry has expressed concerns about the possibility that too stringent rules could hinder innovation.

In conclusion, the AI Act represents a significant step in regulating artificial intelligence, seeking to balance the risks and benefits of this rapidly evolving technology. It remains to be seen how it will be implemented and what effects it will have on the technology sector and society as a whole.

72 views0 comments

Recent Posts

See All


bottom of page