“We have just crossed a point of returning us to a state of total surveillance. The Senate has adopted the use of facial recognition in public space in real time, ”says a user on social networks. A similar message was published by hundreds of people, and relayed several thousand times. It is the “Info Alerts” account that is most frequently taken over.
If for some the news tilts France towards a world at the 1984 by George Orwell, others do not believe it. “It is impossible for such a measure to go so unnoticed”, launches an anonymous person on Twitter, questioning the credibility of the account at the origin of the message. “Elected officials would never have accepted to this extent that we hinder our freedoms”, comments another. What is it really ? 20 minutes explains to you.
FAKE OFF
This Monday, June 12, the Senate did indeed adopt, at first reading, a bill on the subject, with 226 votes in favor and 117 against. But this is to experiment for a period of three years the use of facial recognition in public space. The left opposed it by denouncing “mass surveillance”, and thus managed to have one of its amendments adopted. He adds that the experimentation with facial recognition “does not in any way prejudge sustainability”.
Until now, all biometric data, namely fingerprints or facial recognition, are governed by a very strict framework and are subject to increased legal attention. They can only be processed with the consent of the persons concerned or “on the basis of an important public interest”, as recalled by the National Commission for Data Protection (CNIL). Facial recognition, without the consent of the person concerned, is therefore prohibited and its use must be authorized by law or decree.
A posteriori recognition for terrorism and serious acts
But what does this law adopted by the Senate say? It aims to create a legal framework for experimentation, for three years, of biometric recognition by judicial investigators and intelligence services. This test period will give rise to a government report discussing the possibility of perpetuating or not these uses.
First of all, with regard to the use of biometric recognition a posteriori, that is to say on images already collected and then processed. The text defines its use in judicial investigations for terrorism and serious facts. It will still be necessary that this is based on the images and data already present in the investigation file, and this with the authorization of the prosecutor or the examining magistrate.
Also, the intelligence services will be able, subject to an authorization provided for one month, to use facial recognition on video surveillance recordings in order to find a suspect on the run in a terrorism case. Authorization may also be granted for the identification of the suspect’s entourage.
Real-time usage worries
Then, more sensitively, there is real-time biometric recognition. And this is what scares Internet users. The text this time provides for a use exclusively reserved for terrorism for the intelligence services. Forensic investigators will be able to use it in cases of child abduction or to identify suspects in particularly serious criminal cases. Authorizations will be required for its use. The Prime Minister, the prosecutor or the investigating judge will then give his approval for a period of forty-eight hours.
This real-time facial recognition is not new in France. In February 2019, the city of Nice experimented with facial recognition technology during its carnival by testing several fictitious scenarios, such as identifying fleeing or dangerous people in a crowd.
The objective: to establish a legal framework
The text adopted in the upper house was carried by Senator LR, Marc-Philippe Daubresse and the centrist, Arnaud de Belenet. It follows their report on “facial recognition and its risks”, unanimously adopted by the law commission last year. This “observed a lack of specific legal framework and collective ethical reflection”, indicates the site of the Senate. The objective set in this report was the creation of a legal framework making it possible to regulate practices and “to avoid the deployment of sometimes questionable uses of this highly intrusive technique”.
This is what Marc-Philippe Daubresse recalled from the podium, well aware that the text would not pass without a hitch: “We wanted to collectively define a framework guaranteeing that we would not fall into a surveillance society, it is ie including red lines, and also define a method, a control scheme. »