Law and artificial intelligence

Artificial intelligence is a process that is increasingly used today. In law, we can find a first definition in article 3 of the proposed European regulation of April 21, 2021 of artificial intelligence as being “software that is developed using one or more of the techniques and approaches”, such as “machine learning approaches, those based on logic and knowledge, as well as statistical approaches, “and which can, for a given set of human-defined objectives, generate results such as content, predictions, recommendations or decisions influencing the environments with which it interacts”. But what about artificial intelligence with regard to the law (I), what are the risks of artificial intelligence on fundamental rights (II), and what are the legal advances to limit the risks of the use of AI (III); so many questions that we will try to answer in this article.

Examples of uses of AI in law

AI is a scientific process which will be of interest to several areas of law such as labor law or even personal data law.

Indeed, it becomes a notion of law, because its use emits legal effects. For example, when the question of liability arises following an error or even a malfunction of the algorithm. Or, when it involves intellectual property when it comes to identifying the person with rights to algorithms, just like the area of ​​patent law and copyright law.

In addition, the use of AI is also interfering in the organization of the legal field, whether in the field of business law with contract management and legal research software for example, or even in the field of criminal law with the use of artificial intelligence in court decisions. Or simply in the field of assisted writing, with the now well-known ChatGPT.

We are thus witnessing a real transformation of legal practices brought about by the digital age. As a result, AI becomes both a new source of law and a new tool for applying the law. We are therefore talking about a double impact in the legal world. Thus, artificial intelligence presents itself as both an opportunity and a threat.

Indeed, we believe that in our capacity as a lawyer, we must be careful not to shy away from this new technology and, on the contrary, study ways to use it wisely. However, it should be remembered that whatever the quality of the writings that a tool such as ChatGPT can generate, it will never replace the analysis of a lawyer, especially since it is based on existing data and does not have the power to create reversals of jurisprudence, or even to initiate laws more protective of our fundamental rights.

The negative impact of AI on fundamental rights

In this regard, the increased use of this new technology by the government and businesses makes it necessary to establish a legal framework in order to preserve the internal market, but also fundamental rights.

We note that numerous international bodies have reported the negative impact of AI on fundamental rights, including the National Consultative Commission on Human Rights. They denounce discrimination resulting from algorithmic processing, “remote biometric identification of people in public spaces and places accessible to the public” which violates the right to respect for private life as well as the right to dignity and the right to security. However, we have other examples such as the limitation of freedom of expression and freedom of assembly resulting from the filtering mechanisms put in place by Facebook and YouTube. This can cause reluctance to use AI. Indeed, the development of artificial intelligence could be compared to a regression of fundamental rights. Therefore, establishing a legal framework helps both protect public interests and promote the adoption and development of AI in the internal market.

A legislative embryo

However, currently, this legal framework is non-existent.

There is, nevertheless, a European ethical charter for the use of artificial intelligence in judicial systems and their environment adopted in December 2018.

The latter brings together five fundamental principles in the use of artificial intelligence in judicial systems and their environment. These principles are:

• respect for fundamental rights,
• the principle of non-discrimination,
• that of quality and safety,
• as well as the principle of transparency, neutrality and intellectual integrity
• and finally, the principle of user control.

This charter responds well to the demand from institutions for an AI that does not undermine fundamental rights. However, these are only principles and not obligations, since it is only a charter and not a regulation, a directive or a law; Furthermore, it does not respond to a problem: that concerning disputes caused by artificial intelligence. Finally, as this is a Charter, there is no sanction provided for in the event of non-compliance with these principles.

It is for this reason that we had to go further. Thus, there is a proposed regulation on artificial intelligence presented on April 21, 2021 by the European Commission, called theArtificial Intelligence Act (AIA). The latter aims “to make the Union a leading global player in the development of safe, reliable and ethical artificial intelligence” while guaranteeing “protection of ethical principles”. Therefore, this regulation aims to put in place control of artificial intelligence through requirements, prohibitions, and even sanctions. Thus, through this regulation, we aim to put in place real measures to combat the risks of AI.

It was only on June 14, 2023 that this proposed regulation was adopted by the European Parliament, following a vote in plenary session; This was followed by the start of discussions within the Council on the final form that the law on artificial intelligence will take and its terms of application. However, controversies surrounding generative AI systems that can produce text and images have led to an extension of the text examination time. Therefore, although this is a great step forward, it is estimated that this law will not be implemented before 2026.

Thus, if there is a great involvement of the law in the construction of regulation of the use of AI based on a true balance between legal certainty and respect for fundamental rights and the development and use of artificial intelligence, the law of artificial intelligence remains a right under construction. In any case, the means implemented to control the use of AI show the importance of this process in the legal world. Furthermore, this confirms that law is a changing field, as it adapts to the digital world. The LEGALPROTECH-AVOCATS firm, active in new technology law, is committed to monitoring these aspects.

By Christelle REYNO & Dalhiana PIETRUS