Facial recognition, will it be tested soon in France? Two days later, two laws were passed that make answering that question complicated. On Wednesday, June 14, the European Parliament adopted a draft regulation for artificial intelligence (AI), calling for any use of facial recognition to be banned.
Merger of the Vitaly Card and the ID Card, the risky project of Gabriel Atal
Calendar Chance On Monday, June 12, the Senate voted on a bill by two senators, Marc-Philippe Dupres (LR) and Arnaud de Pellent (Center Alliance). This law, which has yet to be passed by the National Assembly, allows the trial of facial recognition in France for a period of three years, within the framework of the fight against terrorism, and under certain conditions.
sequel after announcement
What does this two-day legislative gap say? Lopes interviewed Noemi Levin, a lawyer at La Quadrature du Net, an association for the defense of digital liberties, in order to better understand the European legislative framework in this area, and the issues raised by the desire for reform in France.
What do you think of the bill adopted by the European Parliament, which opens the law to the regulation of artificial intelligence (AI), and which in particular calls for banning facial recognition systems in public places?
To understand this text, we must go back to the European context of regulating surveillance technologies. This is characterized by two very protective texts: the General Data Protection Regulation (GDPR) and the Police and Judicial Directives, which set out the rules regarding the protection of individuals in relation to the processing of personal data.
The draft adopted on Wednesday comes from a text initially proposed by the committee, dating back to 2021, which envisages the regulation of the artificial intelligence sector. It is more of an economic text than a political one. It contains, for example, obligations for companies wishing to enter this sector.
sequel after announcement
However, Article 5 sets a certain number of red lines, depending on European values, that should not be crossed in the use of new technologies for security purposes. Among these red lines is the prohibition of facial recognition. Released by the commission, the 2021 version made a number of exceptions, weakening the text’s protective framework. One such exception, for example, was allowing facial recognition technology to “track potential victims,” ​​a broad and ambiguous case.
The bill adopted this afternoon is more protective in that it restricts these exceptions. Tout his intérêt est d’interdire la reconnaissance faciale instantanée (lors de grands rassemblements, by example) and his examen va désormais débuter in triologue (entre le Parlement, la commission et le Conseil, which représente les etats membres), ce qui peut prendre several months.
Concretely, what are the risks of using facial recognition processes in terms of civil liberties and individual rights?
Facial recognition is the matching of the face to the civil status. Its use by the state marks the end of anonymity and privacy, as the state gains the ability to know where a person is. However, anonymity is a guarantee of all freedoms: freedom of assembly, demonstration and expression.
sequel after announcement
As a mass technology, facial recognition technology gives itself a way to target already vulnerable demographics in a discriminatory way: immigrants, working-class neighborhoods, activists, and so on. There is no guarantee of their potential use through another diet.
Furthermore, the argument generally put forward for the authorization of these technologies is one of safety or efficiency. However, no study has proven that these techniques are effective, as such. On the contrary, in the countries where it is used, there is a multiplication of false positives, that is, errors in the recognition of profiles. Added to this question of technological feasibility is the question of effectiveness in terms of public policy. Can technology replace human knowledge and fieldwork?
In France, two days before the adoption of this project by the European Parliament, the Senate adopted on Monday a bill allowing experiments to be conducted, for a period of three years, on the use of facial recognition, immediately and a fortiori. Does this law set a precedent?
The chance of success of this law is small, firstly because it must first pass through the National Assembly, and secondly because French legislation is subject to the European framework in this field.
sequel after announcement
However, it indicates a gap between the European framework, not in favor of regulation, and France, the champion of surveillance. In fact, the use of facial recognition is already being used in France, in the criminal record processing file. The TAJ file is particularly relevant to persons involved in certain types of crimes, victims, and persons subject to investigation or instructions to establish the causes of death, serious injury, or disappearance. It contains an image whose technical characteristics allow the use of facial recognition. The bill passed by the Senate goes further, since it authorizes, in addition to this subsequent facial recognition, the possibility of experimenting with instant facial recognition, within the framework of the fight against terrorism.
Should we be afraid of algorithmic video surveillance?
Even if it doesn’t work, it pushes the threshold of what is imaginable in terms of the use of surveillance technologies. This is where this suggestion is dangerous. Especially since it is part of a legislative arsenal that already makes France the spearhead of biotechnological surveillance in Europe. Remember, the law relating to the Olympic Games, passed in May 2023 and certifying the use of video surveillance, is the first of its kind to be adopted in a European country!
