Italy is now officially the first country in the European Union to develop what is established in Regulation (EU) 2024/1689 on artificial intelligence, with legislation that highlights the application of the «anthropocentric» concept.
Although it does not introduce essential modifications to the EU AI Regulation, it does establish some of its own definitions and specifies how AI should be used in such diverse sectors, such as health, labor or public administration.
As explained in a conversation in Confilegal, lawyers who are experts in Italian legislation such as Santiago Martín Aravaca, the concept of «anthropocentric» shows the big difference between the Italian law –DDL or Disegno di Legge– and the draft law of Spain.
On the one hand, Italy focuses on a cultural perspective reminiscent of the Renaissance, a time when humanism It was part of the central axis on which society was articulated, with the objective of «tying» the architecture of the law to the human being and the world.
Meanwhile, The draft law in Spain breathes a guaranteeing ethos and sanctioning, according to Pablo Sáez de Hurtado, multidisciplinary jurist and legal advisor. Although they have forgotten to add sanctions to the Public Administration.
It should be noted that the Italian DDL was approved on September 12although the Spanish draft law may still receive modifications and amendments for subsequent approval.
An «anthropocentric» law in times of AI
The first national AI law has practically focused on a debate very focused on the philosophical field, although establishes several obligations in very current aspects, such as deepfakesthe protection of minors and the use of AI in strategic sectors such as justice.
Regarding the use of deepfakes for malicious purposes – such as impersonating celebrities or political figures – Italy does impose sanctions, since adds penalties of 1 to 5 years in prison in case there is consequential damage.
In this same section, the legislation also toughens penalties for crimes such as fraud or identity theft in cases where AI has caused data leakage illegally.
Additionally, with respect to minors under 14 years of age, they will only be able to access artificial intelligence systems. those who have explicit parental consent for this purpose.
As a great novelty in intellectual property, the Copyright will be protected when using AI for a creation, as long as the human being has actively participated in the creation.
To carry out compliance with the DDL, the regulations mention as governing authorities the Agency for Digital Italy and the National Cybersecurity Agencythe latter to promote the use of artificial intelligence in the security of digital assets.
Spain has another idea for artificial intelligence
Something that draws a lot of attention with a brief look at the Draft Law for the good use and governance of artificial intelligence is that more than half of The regulations focus on infractions, typology and sanctioning elements.
That is to say, unlike the Italian DDL, that of Spain does not focus on teleological reflections on the importance of the human being with respect to technology, but rather on a guarantee standard, to prevent AI from escaping of control.
Thus, it also highlights the need for said sanctioning regime does not imply a high cost for the protection of scientific and technical innovation, with some exceptions and deadlines for the prescription of the facts.
Although it does list in its introduction a large number of reasons why there is a national need to apply the EU Regulation, such as the autonomy that AI systems can achievewhich can «pose a challenge in terms of transparency and traceability.»
It is worth highlighting the backbone importance of the Spanish Agency for Supervision of Artificial Intelligence as a point of contact single with the EU, as well as the competition body in charge of supervision.
In short, Italy and Spain have a very different visionfrom the conceptual framework, to complete the regulation of AI.