The European Union reaches an agreement to regulate Artificial Intelligence

The European Union reaches an agreement to regulate Artificial Intelligence

The European Union (hereinafter EU) has taken a leadership position in the race to regulate Artificial Intelligence (hereinafter AI), as on December 8th, 2023, all member states met for 3 days and reached a provisional agreement regarding the regulation of AI, a project known as the Artificial Intelligence Act. The European Commissioner for Internal Market and Services, Mr. Thierry Breton, has described it as “a historic agreement”.

This provisional agreement seeks to achieve the main objective of protecting citizens from the risks associated with this constantly developing technology. Furthermore, it aims to encourage the use in the European market of safe AI that respects the fundamental rights and values of the EU.

The most important aspects that have been addressed in the provisional agreement, which still has to be ratified, are the following:

Firstly, the use of real-time surveillance and biometric technologies, including emotional recognition, has been banned. Only three exceptions have been established where the police could use this technology: (1) in the event of an unexpected terrorist attack threat; (2) the need to search for victims; and (3) the pursuit of serious crimes.

Moreover, limits have been established regarding generative AI, that is, the one used to create texts, music, and audiovisual content. In this sense, there will be an obligation to inform when the content being reproduced, published, or disseminated is the work of AI.

Furthermore, direct bans have been established in relation to certain activities that could affect fundamental rights:

→ Systems that use biometric identification of people based on their political, religious, philosophical beliefs, race, sex, or sexual orientation cannot be used.

→ Audiovisual material from Google, Social Networks, TV, etc., cannot be used to create a database for facial recognition.

→ Abusive emotion recognition systems at work or schools are not allowed.

→ Systems that can manipulate people’s behavior are not permitted.

→ Systems known as “social scoring,” used to socially score people, are not allowed.


Likewise, it has been decided to create an Office of Artificial Intelligence, which will be responsible for supervising AI systems that enter the European market, contributing to the norms that may arise in this regard, and ensuring that existing norms are complied with by all member states.

In parallel, there will be an Artificial Intelligence Committee formed by all member states that will be responsible for coordination and consultations that may arise.

To support this Committee, a Consultative Forum will also be created, which will serve for all those companies, SMEs, societies, and figures from the academic world to have the appropriate technical knowledge that the Committee can provide.


Considering all the above, obviously, fines have been established for breaches of the AI Act, setting a percentage of the global annual turnover of the offending company in the previous financial year or a predetermined amount if this were higher. The main penalties are as follows:

→ 35 million euros or 7% of the turnover for infringements related to AI applications.

→ 15 million euros or 3% of the turnover for non-compliance with the obligations established in the AI Act.

→ 7.5 million euros or 1.5% of the turnover for providing inaccurate information.

However, the provisional agreement sets more proportional limits to the administrative fines that can be imposed on SMEs and startups in case of infringement.

Regarding the entry into force, the provisional agreement determines that the AI Act should apply two years after its entry into force, with certain exceptions for specific provisions.

Therefore, we are facing a Regulation project whose main purpose will be to ensure that AI systems introduced into the European market are safe, as well as to stimulate investment and innovation in the field of AI in Europe, effectively addressing a global challenge in a technological environment that is experiencing rapid evolution and in a key sector for the future of our societies and economies.

Otros artículos de interés

Social Media and Data Protection

The relevance of social media is, today, beyond question. It is a service of the information society, the provision of which involves data processing, whether to[...]

Leer más