April 8, 2022

By Nicolas Hénin, Researcher at EU DisinfoLab

The last French presidential election five years ago was marked by the Macron Leaks and the spectacular dissemination of data, massive in quantity even if most of it was fairly innocuous, hacked from the candidate’s campaign inner circle. Even before the outbreak of the war in Ukraine, which required a reinforced posture of vigilance against cyber or information attacks, securing the 2022 presidential election was a priority, for state actors, civil society, and platforms alike.

Government-led initiatives

The 2018 law “on the fight against the manipulation of information”

This law is arguably the first response to the Macron Leaks and the consideration of threats of electoral interference.

Two previous legal texts regulate these issues: 

  • The 29 July 1881 law on the freedom of the press provides (Article 27) for a criminal offence punishable by a fine of 45,000 euros for the “publication, dissemination or reproduction, by any means whatsoever, of false news, fabricated, falsified or falsely attributed to third parties when done in bad faith, it will have disturbed public order or will have been likely to do so.”
  • Article L97 of the Electoral Code also makes it possible to prosecute someone for spreading false news: “Those who, by means of false news, slanderous rumours or other fraudulent manoeuvres, have surprised or diverted votes, determined one or more voters to abstain from voting, will be punished by one year’s imprisonment and a fine of 15,000 euros.”

The 2018 law brings several novelties:

  1. Legal action in summary proceedings

During the three months preceding a national election (presidential election, legislative election, senatorial election, elections to the European Parliament, and referendum), a summary judicial action is possible to rapidly interrupt the dissemination of a publication – according to the following criteria:

  • The false news must be obvious,
  • It must be massively and artificially disseminated,
  • It must lead to a disturbance of the public peace or the integrity of an election.

The interim relief judge then has 48 hours to rule on the nature of the information and possibly order its depublication.

  1. Duty of cooperation for platforms

Outside election periods, the law creates a duty of cooperation for platforms, which must implement measures to combat the dissemination of false information likely to disturb public order. They must put in place an easily accessible and visible mechanism for internet users to report false information. Platforms must also designate a legal representative in France to act as a contact person. Some of these provisions are taken into account by the Digital Services Act (DSA).

  1. New powers for the French audiovisual regulator ARCOM

The Authority for the Regulation of Audiovisual and Numeric Communication (ARCOM) resulted from the merger between the Superior Audiovisual Council (CSA) and the copyright-focused High Authority for the Dissemination of Works and the Protection of Rights on the Internet (Hadopi). The regulatory authority can now suspend the broadcasting of a television channel that is proved to be “under the influence” of a foreign state, in case it deliberately propagates false information that could potentially affect the integrity of the vote.

Viginum

The Vigilance and Protection Service against Foreign Digital Interference, known as Viginum, is a national agency set up in France to combat information manipulation. This service with national competence, attached to the General Secretariat for Defence and National Security (SGDSN), has been operational since 15 October 2021.

“Our objective is to be able to detect attacks when they come from abroad, to be able to characterize them and thus be able to attribute them in a certain way,” according to Stéphane Bouillon, Secretary-General of the SGDSN. In concrete terms, Viginum’s role is to monitor, detect and analyse the operations and techniques used by foreign actors to disseminate and amplify content hostile to France online with the aim of damaging the nation’s interests. This is the case, for example, of attempts to manipulate opinion on social networks using bots (automated accounts).

Viginum, headed by former Court of Auditors magistrate Gabriel Ferriol, has an annual budget of 12 million euros. It is planned to have 65 employees, mainly analysts, data engineers, and digital media experts. These agents work with open sources information (media, Internet, social networks, professional and government publications, etc.).

Viginum’s management of personal data is supervised by the CNIL (French National Commission for Information Technology and Civil Liberties), and subject to the control of an ethical and scientific committee, chaired by the Council of State Béatrice Bourgeois-Marcheau.

Once Viginum identifies a threat, internal assessments are made on the responses, for example, deciding to make such interference public or to “act with the social networks so that they intervene and delete the disputed accounts”, according to the French ambassador for digital affairs Henri Verdier.

Raising awareness among political parties

As it had been done for the previous elections, in September 2021, the Prime Minister gathered all his ministers at the International Cyber Security Forum (FIC). In parallel, the ANSSI (National Agency for Information Systems Security) brought together the cybersecurity managers of the candidates’ campaign staffs to provide them with its recommendations. 

Subsequently, ARCOM holds weekly meetings with the political parties to monitor disinformation and interference.

Initiatives led by civil society

GVNIE Consortium

A consortium of researchers, the Digital Vigilance and Electoral Integrity Group (GVNIE), has been formed to anticipate harmful actions affecting the integrity of the vote in order to better understand them. It is part of a program developed by Reset and the Luminate Foundation to monitor digital threats to electoral integrity. 

The objectives of this group are to:

  • Pool and optimise existing monitoring efforts, through the comparison of different observation angles and methodologies;
  • Alert platforms to potential actions that could impact the integrity of the election;
  • Engage in a common reflection on the large-scale circulation of content that may constitute threats to society;
  • Raise public awareness of research results and make them accessible to all;
  • Work towards a better understanding of these phenomena by the public authorities and greater accountability of digital platforms, especially in the context of the forthcoming application of the European DSA.

The member organizations of the group are:

2022 French Election Dashboard

A member of this consortium, the Alliance for Securing Democracy housed by the transatlantic advocacy organization the German Marshall Fund, has also launched, with ISD’s assistance, a platform on its website, the 2022 French Election Dashboard. It provides a summary analysis of the topics and themes promoted by foreign and domestic actors during the 2022 French presidential election. The dashboard tracks the outputs of Russian and Chinese government officials and state-owned media (the ban of the Kremlin media has de facto reduced the scope of this portal), as well as other foreign state-owned media outlets targeting French audiences on Twitter, YouTube, Facebook, and state media websites.

22vlalapub!

CheckFirst, a company that provides tools and expertise to those fighting disinformation, has launched the 22vlalapub! website, a device for monitoring and analysing ads about candidates, their parties and campaign themes related to the presidential election on Meta platforms. It uses the Facebook Ad Library, and ads considered to be “about a social, electoral or political issue” purchased for distribution on the group’s platforms (Facebook, Instagram, Messenger) and mentioning the candidates, their political entity, or a list of campaign themes that are at the heart of public debate. For each candidate,  22vlalapub.fr aggregates the ads and associated data disseminated through their personal page and through their party page, the ads disseminated by other stakeholders that mention the candidate’s surname, and finally, those that mention the candidate’s party. 22vlalapub.fr also aggregates keywords associated with recurring campaign issues.

Initiatives by platforms

Google & AFP

In December 2021, AFP and Google France set up a program to fight disinformation entitled “Objectif Désinfox“. This training and collaboration program, aimed at French newsrooms, is based on an alliance of media and fact-checking organizations interested in countering disinformation. Although this project is not directly linked to the upcoming presidential election, the election has certainly inspired it.

The project is intended to assist the partner newsrooms in their research, monitoring, and verification mechanisms.

The main objectives announced in the launch press release are:

  • Free online thematic training courses from mid-January 2022 for all member media and fact-checking organizations of the alliance, in order to train their journalists in digital investigation (digital verification tools, improvement of monitoring processes, fight against cyberbullying, etc.).
  • The creation of a collaborative platform for exchange between alliance members, which will include a system for reporting false or misleading content to the general public and which will feed all partner newsrooms.
  • Monthly thematic meetings led by AFP, to which all members will be invited, with the participation of external speakers and experts (researchers, sociologists, etc.).
  • The provision of AFP’s political fact-checking content for alliance members on the AFP Factuel website and on the “AFP Fact Check” YouTube channel.
  • A program of free assistance and support for fact-checking newsrooms that are members of the Alliance in the creation and production of their own content. This content will be labelled with the Objective Disinfox label, and AFP will produce videos of the best productions.

Ironically, one of the first public announcements related to this project, a job advertisement, was targeted by a violent raid by the far-right on social media for mentioning the search for a journalist “of diverse background”.

Google & the 4Rs

Google highlights the “4Rs” strategy already in place on YouTube

  • Removing videos that violate the law or the terms of use, 
  • Reducing the visibility of content that is borderline illegal, 
  • Relaying content from verified sources, 
  • Rewarding (by paying) creators of quality videos.

Google tracks major attack attempts in a quarterly report on its website – and also publishes a transparency report on government requests to remove content

Meta

The group that controls Facebook, WhatsApp and Instagram, has announced the opening of a “virtual operations center” for permanent monitoring and a collaboration with the French media. WhatsApp users can, for example, contact Agence France-Presse directly (using the number +33 6 47 08 70 46) to report false information, which the news agency can verify. After the outbreak of the war in Ukraine, this service seems to have been in great demand. Applicants received a message warning them of longer processing times.

In addition, a series of media education campaigns developed in partnership with AFP, Loopsider, and France 24 were to be deployed on Meta’s platforms.

On Instagram, Meta launched an account called “Fake ou quoi” with the media company Loopsider, and a series of videos designed with Konbini called “Ça vote”. On Instagram stories, or in Facebook news feeds, calls to register to vote while possible, or to go and vote when the time comes. Meta publishes periodic reports on its takedowns for Coordinated inauthentic behaviours (CIB) or on its actions to ensure election integrity, although there has been no specific focus on the upcoming French elections.

Twitter

Tweets containing “false or misleading information about how to participate in the election” will be removed. However, the platform has another measure for posts challenging the outcome of the presidential election: it will put a “label” on messages that “undermine confidence in the election or its outcome”.

The social network will also include contextual banners about the election, which will be based on “official, trusted sources” and fact-checkers.

However, there will be no extension to France of the Birdwatch feature, which is being tested in the United States. This collaborative tool, launched in 2021, allows participants to add fact-checking notes to tweets they find misleading. Twitter publishes periodic reports on “​​inauthentic influence campaigns”, but also in this case no content was directly related to the French elections.

TikTok

The Chinese video-sharing and social networking mobile app has rolled out a 2022 Presidential Election Guide available under the app’s “discover” tab. According to a press release posted on its website, “this guide provides users with reliable information from official sources on the electoral process: how to register to vote, when to vote, where to vote, how to vote and how to vote by proxy, as well as the list of all candidates (source: Constitutional Council). It will be progressively updated and will include educational videos from the French government’s TikTok account.”

The guide will also feature links to the French government’s official election website and to the website of its partner Génération numérique, an association specialized in media literacy.

TikTok considers itself not to be a place for political expression and does not allow political advertising.

To prevent the spread of false information, the platform collaborates with fact-checkers, including AFP. When the platform identifies a piece of false information, a label is applied to it. If a user tries to share the flagged video, he will receive a message reminding them that the video has been flagged for unverified content. This extra step invites the user to think and reconsider their next action, before deciding to “delete” or “share anyway”. These videos cannot be recommended in the “for you” feed.