Reading time: 18 minutes
Translation by AB – April 20, 2020
Anti-fake news law
After several months of debate, the Parliament adopted on November 20, 2018 the bill relating to the fight against the manipulation of information dubbed “anti-fake news law”. Emmanuel Macron promulgated it on December 20, 2018 after a final seizure of the Constitutional Council by concerned deputies and senators. Concerned? Reading the few paragraphs of the bill1, we understand why.
The “anti-fake news” law aims first to fight against the pernicious effects of false information during election periods. Thus, platform operators, who already have certain transparency and loyalty obligations (Article L111-7 of the French Consumer Code), must fulfill additional obligations in the pre-election period2. In particular, they must provide the identity of those who promote, in return for payment, information content related to a “debate of general interest”. Well, why not. But the new article L163-2 specifies that, during this same period, the summary judge [in French: juge des référés] has 48 hours to eventually stop the dissemination of “inaccurate or misleading allegations or accusations of a fact likely to alter the sincerity of the upcoming vote [which] are deliberately, artificially or automatically and massively disseminated through an online public communication service”. However, the Electoral Code already provides for the punishment of:
Surprising or hijacking voters or leading voters to abstain, using “false news, slanderous rumors or other fraudulent tactics”.
But we can see the essential difference and the beginning of the problem: the anti-fake news law provides for the a priori intervention of a judge on the basis of a presumption of deception or inaccuracy. It is obviously a question of putting an end to the dissemination of false information as quickly as possible. It is this objective, a priori praiseworthy, which is flawed. First, there is this evidence noted by the Senators themselves3 :
How could the summary judge, in 48 hours, establish the alteration of an election which has not yet taken place?
These same Senators also consider the text disproportionate since it would prevent the disclosure of allegations “that would only be misleading, but not inaccurate, even though they may be part of the democratic debate“. It would therefore attack the principle of freedom of expression and the sensitive issue of “official truth”. We shall return to this essential point.
[ The “CSA” – Conseil Supérieur de l’Audiovisuel – is the French authority which mission is to “guarantee the exercise of freedom of audiovisual communication in France” ]
The foreign media are also in the cross hairs, those who would be wary of manipulating opinion during the pre-electoral period. This time it is a question of adjusting the so-called “Léotard” law on freedom of communication by giving the CSA the possibility of suspending the (electronic) broadcasting of any media controlled “by a foreign State or placed under the influence of that State [which] deliberately disseminates false information likely to distort the truthfulness of the ballot”. It is even possible to terminate the agreement with this medium if the case is serious, i.e. “if the service which is the subject of the said agreement is detrimental to the fundamental interests of the Nation, including the regular functioning of its institutions, in particular through the dissemination of false information”. Of course, one thinks of Russia, accused of having rigged, among other things, the 2016 American elections.
The CSA therefore has in principle the same kind of prerogatives with regard to foreign media as the summary judge with regard to platform operators. The difficulties are therefore of the same order but with an additional suspicion: what about the real independence of the CSA?
Finally, the CSA must contribute to the fight against the dissemination of false information by controlling the transparency obligations made to platform operators concerning their algorithms, the use of their users’ personal data, the origin of the content offered, the promotion of content from news agencies, education, etc.
The press in a vise
Truth management concerns first and foremost those who produce and control the “official truth”: political actors and bodies, the judiciary and the media. Of course, the latter denounce a text that contradicts an essential principle of the 1881 law on the freedom of the press4 :
There is no a priori control (censorship) but only a posteriori, and by judicial means, which offers guarantees against arbitrariness.
However, what is likely to “alter the sincerity of the upcoming vote” is open to a priori interpretation and is therefore potentially arbitrary.
From now on, the State will be involved in defining what is true and what is false. And then there are the increased powers of the CSA, whose old dream was to see its functions extended to the Internet. Here is a modest, but real, embryo of a censorship commission.
The press wishes of course to respect the spirit of the law on freedom of the press and to report only to its readers, if necessary to the courts. However, in order to “fight” the GAFA, which has the power of a State, only a State can, which therefore imprudently interferes in trying to unravel the true from the false. In this face-to-face confrontation, the traditional press is caught in a vise.
Justice in doubt
The legal world also doubts the interest and effectiveness of this law. The arguments are quite similar to those of Senators. They also add the possibility of using this text for delaying tactics5 :
Imagine that François Fillon [a former French Prime Minister candidate from the Presidency] used this text during the revelations about possible misappropriation of public funds. He could have asked the summary judge to order “Le Canard enchaîné” [a famous satirical French Newspaper] to stop his investigations on the pretext that they are likely to alter the sincerity of the vote!
And they also point to deeper evidence about the nature and real effects of misinformation:
Experience shows that the logics of influence favoring extreme right-wing ideas – migrants, burkini, etc. – are implemented two years before an election, whether American, British or German.
This anti-fake news law is therefore not obvious in its application and carries multiple possibilities of diversion. This is indeed a sign of a difficulty, both in grasping the phenomenon at the right level and in assessing its real effects concerning public order. How did we end up having to design such a questionable text? Was there an urgency to legislate? Why is false information such a problem today?
Propaganda and false truths have always existed, but today we are confronted with a phenomenon on an unprecedented scale and also apparently paradoxical: how is it that so much false information, alternative facts, etc. can flourish when the Internet offers everyone so many means of investigation and verification?
It is well known that “truth” is established by consensus rather than by evidence. It is a majority opinion about “reality”. However, the digital world is a powerful catalyst of coalescence, of domains of intersubjectivity (groups, forums, blogs, platforms…) emerging and disappearing at an unprecedented speed. The “democratic truth”, patiently elaborated, defended by the usual democratic organizations (press, justice, elections…), resists it badly. The number and speed of circulation of the alternative truths are easily observable characteristics. But there is also their astonishing robustness, illustrated by two famous examples: the election of Donald Trump and the supporters of flat earth. This is an opportunity to receive at least two lessons.
Lesson # 1: despite the Internet, everything is potentially fake
Donald Trump’s inaugural ceremony on January 20, 2017 was attended by all the American and international press. 250,000 people gathered in the National Mall, a relatively sparse crowd compared to the million people who came in 2009 to attend the inauguration of Barack Obama:
At the time, some people doubted the journalistic interest of these pictures. But the so-called “mainstream” press was blowing the whistle on Donald Trump and was conducting a permanent lawsuit for illegitimacy. These photos showing a “badly elected” President were therefore circulating. This type of photo is reminiscent of a montage, a “fake”. But in spite of everything, these are photos, irrefutable elements, aren’t they? But, on January 21, 2017, White House spokesman Sean Spicer held his first press conference and…6
… he accused the media of deliberately underestimating the size of the crowd for President Trump’s inaugural ceremony and stated that the ceremony had drawn the “largest audience to ever witness an inauguration – period – both in person and around the globe”
And he used some odd arguments. Nevertheless, the next day, in another meeting with the press, reporter Chuck Todd (NBC News) asked Kellyanne Conway to explain why Spicer used a blatant lie. She replied: “Don’t be so overly dramatic about it, Chuck, you say it’s a lie … our White House press secretary, Sean Spicer, was merely presenting alternative facts“, namely that the crowd at the Trump nomination could not be proven or quantified:
Chuck Todd upbraided Kellyanne Conway with the claim that “alternative facts are not facts; they’re falsehoods”.
It was amazing. We were witnessing a state lie live in front of the press, contradicted by photographs and video available on the Internet, and journalist Dan Rather, a recognized actor in “democratic truth”, almost ran out of words to describe this situation, to the point of invoking Orwell in a famous post:
These are not normal times. These are extraordinary times. When you have a spokesperson for the president of the United States wrap up a lie in the Orwellian phrase “alternative facts” […] Facts and the truth are not partisan. They are the bedrock of our democracy. And you are either with them, with us, with our Constitution, our history, and the future of our nation, or you are against it. Everyone must answer that question.
This episode revealed the place of truth in 2016, slowly undermined by years of almost invisible communication on social networks, which have become, unbeknownst to a press that saw little or nothing coming or wanted to know about it, the place where opinion is formed. Donald Trump was easy saying to this press: “You are fake news“! The trap had closed.
Lesson # 2: In a meaningless world, any truth is good to take
The case of the supporters of a flat earth, the “flat-earthers”, exposes us to a sort of maximum vertigo7:
The flat-earthers pose a kind of intellectual challenge: it’s so overwhelming that it’s hard to believe that they believe it.
Moreover, it is likely that some flat-earthers are almost indifferent to the reality of a flat earth and only motivated by the intellectual challenge of defending it (the game is a strong motivation for some fake news propagators). James Finn answers on Quora to the question “Have you ever met a flat-farther who finally converted to globe earth? What made him/her change his/her mind?“8 :
I never convinced a single one of them of anything. They aren’t interested, typically, in evidence. Their belief system is religious in nature in that it’s largely irrational and based primarily on faith.
Do not imagine that scientific arguments and photos can overcome the faith in a flat-earth. Let us take an example. On the occasion of Felix Baumgartner’s historic jump, on October 14, 2012, many photos taken by his team circulated and tickled the flat world, like this one:
Flat-earthers have different explanations for the curvature behind Baumgartner: either it’s the “fish eye” cameras that distort the horizon (obviously flat), or it’s photoshopped, or Baumgartner never did that… A photo, like those of Donald Trump’s investiture, “proofs”, “evidences”, nothing helps.
Finally, for rational thinking enthusiasts, there is this useful argument that often miss: releasing stress! We are approaching the source of robustness of alternative truths, this form of “joy” provided by a framework of free truth, whether virtual or real, whether it is a digital group, a sect or a political party to which we freely choose to join9 :
The event [“International Conference on Flat Earth” in 2017, in North Carolina] proceeded like any classic conference – exhibitors in the hallway, speakers in the auditorium – with breaks to stretch your legs and a bar open in the evening. In the corridor, a white-haired craftsman showed splendid wooden models of flat earth to the participants, while old ladies summarized conspiratorial books self-edited to their girlfriends, drinking tea. […] On an individual level, the absolute denial of “consensual” reality provides flat-earthers with a means of confronting a complex and violent world. … As we chatted under a sky riddled with chemtrails, bombarded by the 5G waves sweeping across the parking lot, Watsun Atkinsun confessed to me that he agreed with this idea: for him and his companions, flat earth is a form of liberation from the stress that the world exerts on us. It gives us the opportunity to reinterpret reality in a more pleasant, more serene way.
Even so, some flat-earthers have ended up abandoning their conviction, but not under the weight of scientific evidence. It’s rather unexpected:
Barry eventually came around, though. He tells me he was smoking a giant doob one night when the realization hit him that all the leading flat-earth YouTubers are making big bucks from their vids. So, it must be a scam. He decided not to believe it anymore.
Truth works in mysterious ways!
Propagation: there’s no stopping the fake news
These forms of truth are like forms of life which, with the digital world, have found an extraordinary biotope. Indeed, this environment is purely informational, with very few counterforces from “reality”. All information, true, false or whatever, is created equally and survives by visibility, i.e. by propagation. Thus, fake news look tremendously like “true” news. The means available to everyone (social networks, editorial tools, etc.) make it possible to publish professional-looking content easily and inexpensively, to modify photos and videos (AI and its shadows) and to reach a large audience. Newspapers no longer have a monopoly on credible forms (“In effect, we are all CNN now” wrote the journalist Mathew Ingram in 2015).
Thus, the “pay-per-click” economic model urges the press to offer sensational, in any case sensational titles (the themes of AI and neuroscience are major providers) and words likely to optimize natural referencing by search engines. The equivalent of neon signs in a street saturated with shops … Then, how to direct ourselves in this mixture of real “sensationalized” information and fake (and of course) sensational news? Trusted third parties have gradually disappeared. Thanks to social networks and platform operators, no one comes between us and our desires, no one normalizes our relationship to “reality”. Everyone is left to their own devices and satisfies their need for meaning and coherence in their local, often digital, group (sometimes echoing reality on a roundabout or in a conference of flat-earthers …).
Is there any coherence, any general line, in these cumulative reasons for the proliferation of fake news in this new biotope? Perhaps a quest for meaning, emotion and “being”, in a (seemingly) hyper-rational world …
Without truth, no Nation
Obviously, the guarantors of the official truth have an interest in reacting and they are the only ones: they are the press, political powers and justice.
It is not a question of giving in to conspiracy or relativism. In a democracy, official truth is a common good, just as territory is a common good. As Dan Rather said, it is “the bedrock of our democracy“. He goes further by saying that it constitutes us. Thus, French history is the history of French truths. It is for this reason alone that “the earth is flat” must be collectively considered as a fake news. False information, as long as it has the power to harm the common truth, must therefore be rejected as a foreign body. But by what means (aside from a painful law of circumstance)?
Truth is expensive to seek for and the value is not catched by those who discover it but by those who spread it, like Google News, Apple News or Microsoft News… There is therefore first the question of an “economy of truth”. Who finances the production of a democratically acceptable truth? Are we citizens still willing to pay for a “trusted third party” of truth, whether it is a public or private media, for example, for a “fact checking” press? Is fact checking a good solution?
The trumpist, the flat-farther, the evangelist… will not go to a fact checking site (except maybe to check that we are interested in what they think…). The fact-checking will only be used by people who have been trained to question themselves, that is to say quite a few people. There is also the mass (number / speed) of facts to check. A choice must therefore be made which can itself be interpreted as manipulation or political orientation. The most enraged can easily become immune, like a well-known American right-wing fundamentalist site with a specific rhetoric of the “fake” easy to recognize:
Beware of Snopes. I do not really think they are lying, but they present Conservative emails in a very unfavorable way. So unfavorable that it’s hard to believe their work is objective.
On a more moderate note, Daniel Schneidermann of the French newspaper “Libération” wrote about the Decodex (a fact-checking website) of the rival newspaper “Le Monde”10 :
It’s like asking a taxi company to label Uber or asking real estate agencies to say if Airbnb is a cool app. A newspaper favorable to globalization, Le Monde classifies pro-globalization newspapers in green and the others in red. Le Monde is purely and simply in conflict of interest. Judge and party11.
Fact checking is political, therefore suspect, even between peers …
Tackling social media?
Is it a good idea to tackle social networks, the natural and original environment of fake news?
Asking social networks to participate in the regulation of the truth of contents (and not their legality) is to propose what we regularly denounce here in terms of ethics: asking those who are responsible to propose solutions to us. Nicholas Lemann of The New Yorker writes12 :
It’s a sign of our anti-government times that the solution proposed most often is that Facebook should regulate it. Think about what that means: one relatively new private company, which isn’t in journalism, has become the dominant provider of journalism to the public, and the only way people can think of to address what they see as a terrifying crisis in politics and public life is to ask the company’s billionaire C.E.O. to fix it.
But since everyone is suspicious on this subject, we see at the same time the flaw in his statement: Facebook does not fit in the heart of The New Yorker.
And if we sometimes give the “biotope” a chance (before we legislate)?
There are self-regulatory forces that we do not allow time to prove themselves. Social networks have a greater interest than anyone else in fighting this phenomenon to preserve their business model: advertising. Advertisers don’t like being associated with fake news that can put their brand in trouble.
This is why the “billionaire” mentioned by Nicholas Lemann already has the solution of algorithms for his business13. The truthfulness of the information being only very relative and not really the subject, it is rather a question of algorithms aiming at the dynamics of digital propaganda. Thus, one will notice that a fake news is often accompanied by a retouched image, a message unrelated to the illustration, a text with an approximate language, a “biased” speech followed by numerous exclamation marks, etc. Fake news also has a very particular way of spreading. AI algorithms (“Geometric Deep Learning”, etc.) are able to identify a large number of these informational trolls in this way.
This digital biotope is much better understood by young people than by the generations that have known pre-digital history14:
Internet users over the age of 65 tend to share more false information on Facebook than younger users, regardless of their education, gender, race, income or the number of links they share. Age therefore appears to be the criterion that predicts their behaviour better than any other characteristic, including party membership.
Even if the new generation is better prepared, education must be a top priority, especially in the handling of digital information. As we have just seen, this information is of a different nature from paper, radio or television information. New teaching must therefore be designed. Some initiatives are outstanding. For example, in France, the teacher Rose-Marie Farinella has specialized in fake news education15. She has developed a full educational program, 45 minutes a week throughout the school year. The children learn to read and write digital information, to stop at every sensational piece of information, to cross-reference sources, etc. The whole work is exciting and reminds us that a bright child is curious, responsible, relevant and can be prepared to live in a legal framework that respects him/her:
In the classroom, one student said that it would take millions of years for humans to control all the content on the Internet. Another replied that it could only be the work of robots and algorithms. Another pointed out that the advantage of an algorithm or a robot is that they have no religion, no bias …
Pas mal pour des élèves de CM2…
Let us finally mention the work of Christophe Michel, a teacher close to Rose-Marie Farinella who declared before the European Parliament:
Freedom of expression is not only about defending the freedom of everyone to defend reasonable, fair, correct and true information. To be for freedom of expression is to defend the right of everyone to provide information that is unreasonable, false, incorrect, unfair, even shocking or disrespectful.
But knowing the difference between the two is essential and this should be the role of education first, not legislation.
Back from the “bush”
First, in the digital environment, fake news is not only information but rather counts for the immediate shot of emotion it provokes (remember flat-earthers). Fake news is a quasi-symbol of rallying without any real connection to reality.
Second, the authors of the French anti-fake news law may have wanted to believe that Donald Trump was elected because the Russians had taken control of social networks. But maybe a) Hillary Clinton was generally disliked, b) many Americans were really fed up with her, c) Donald Trump is a powerful meme, etc.
Third, Presidential or European elections can perfectly accommodate fake news. It is not because of them that these elections could lead to a result that would not suit the instigators of the law… In his country, invisible to the authorities, the flat-earther is happy and already knows who he will vote for.
We therefore respectfully suggest to the instigators of the law a visit to Rose-Marie Farinella, where they will be able to get close to youthful intelligence, to (re)take a taste for educational solutions, those that allow the preservation of a common trust without which no freedom is possible.
1. ↑ On the French Parliament website – Proposition de loi relative à la lutte contre la manipulation de l’information
2. ↑ 3 months before the election; for the European elections of May 26, 2019, these obligations have therefore applied since February 1.
3. ↑ Marc Rees / Next Inpact – November 21, 2018 – Loi contre les « fake news » : à peine votée, déjà attaquée devant le Conseil Constitutionnel
4. ↑ Etienne Gernelle / Le Point – January 31, 2019 – Les Gafa (et l’État) contre la liberté de la presse
5. ↑ Laurence Neuer / Le Point – July 2, 2018 – Loi anti-fake news : les juristes tirent la sonnette d’alarme
6. ↑ Wikipédia – Faits alternatifs
7. ↑ Alice Maruani / Rue 89 – August 20, 2016 – J’ai essayé d’interviewer quelqu’un qui croit que la Terre est plate
8. ↑ James Finn / Quora – Have you ever met a flat-Earther who finally converted to globe Earth? What made him/her change his/her mind?
9. ↑ Rajiv Golla / Vice – January 4, 2018 – Deux jours avec les partisans de la théorie de la Terre plate
10. ↑ Daniel Schneidermann / Libération – February 5, 2017 – Décodex décodé
11. ↑ Which will not prevent the same Daniel Schneidermann from judging “fact checking powerless but necessary”! (Le fact-checking, impuissant mais nécessaire)
12. ↑ Nicholas Lemann / The New Yorker – November 30, 2016 – Solving the Problem of Fake News
13. ↑ Laure Aventini / Yubigeek – October 5, 2017 – Des algorithmes pour traquer les “fake news”
14. ↑ Anaïs Cherif / La Tribune – January 10, 2019 – Sur Facebook, les personnes âgées partagent plus de fake news que les jeunes
15. ↑ Émilie Brouze / Rue 89 – January 9, 2017 – « Hoax » : l’instit qui apprend à ses élèves à devenir des détectives du Web
16. ↑ YouTube – Hygiène Mentale