Facebook: "The fight against online extremism should not come down to platforms, but to the legislator"

Consultant and political analyst, Mathieu Slama works for several media.

He has published

The War of the Worlds, Reflections on Putin's Crusade Against the West

, (ed. De Fallois, 2016).

To discover

• Michel Houellebecq: "A civilization which legalizes euthanasia loses all rights to respect"

The most worrying drifts are often inspired by the best intentions. We learned, in an article by the American media CNN, that Facebook had launched a test operation in the United States to support its program to fight against online extremism. As part of this operation, American users of the platform received an alert that read the following message: "

Do you think that a friend of yours is becoming extremist?"

", Before specifying:"

We are concerned about the fight against extremism at Facebook. Other users in your situation have received help from us

".

If you click on the alert, it redirects the user to a page where he can specify the information he has which is then processed by Facebook.

This test program could appear anecdotal if it did not form part of a fundamental trend which is imposing, little by little, in the name of the fight against extremism and hate speech, a digital world of surveillance, censorship and of denouncement.

Facebook, like other platforms, does not act spontaneously. Long accused of favoring a lax approach to hate speech expressed on his platform, Mark Zuckerberg is changing his general philosophy and approaching the practices of the social network Twitter, which has chosen for several years a very interventionist line in the regulation of content. And where Trump was a staunch supporter of free speech on platforms, going so far as to threaten them in the event of excessive violations of freedom of speech, it is different with the Biden administration which defends one line. much more coercive and disciplinary in terms of regulation. So there is a context behind this latest innovation from Facebook,initially motivated by the desire to show white paw in the face of political pressure which is becoming stronger and stronger. And let's not forget, it's easier for platforms to regulate more than it is to pay more taxes. The demonstrations zealous in terms of regulation are primarily there to calm the political storm and to give themselves a better image at little cost.

In France, the Avia law of June 24, 2020 provided for draconian regulatory measures, obliging the platforms to censor any hate speech themselves within 24 hours, without any judicial intervention and at the sole discretion of the platforms.

Mathieu slama

For a year now, we have witnessed a worrying tightening of online regulatory measures. Everyone remembers the spectacular decision by Facebook and Twitter to ban Donald Trump, then still sitting president, after the events on Capitol Hill last January. This decision, by its scope and gravity, constituted an extremely serious precedent: it was now possible, for the GAFA, not only to censor a head of state in office, but also to deny him any possibility of expression by removing him. purely and simply its official accounts. Note that a month ago, Facebook confirmed the deletion of Trump's account for a period of two months, announcing at the same time new measures to fight against online violence,especially with regard to policies which are now subject to the same regulations as ordinary users.

In France, the Avia law of June 24, 2020 provided for draconian regulatory measures, obliging the platforms to censor any hate speech themselves within 24 hours, without any judicial intervention and at the sole discretion of the platforms, therefore, according to criteria particularly vague (when does hate speech start? where does it end?). The Constitutional Council fortunately censored this provision, but the door was now ajar to a major tightening of the regulation of social networks in France.

Other initiatives are emerging and are just as worrying. Since 2020, Twitter has been performing tests with the help of artificial intelligence to alert each user of the potentially offensive nature of a draft tweet containing "insults" or "hateful remarks". Thus, when you want to publish your tweet, an alert is displayed to warn the user that the content is potentially offensive, inviting them to rewrite it. It is easy to understand the novelty of Twitter's approach, which aims to act a priori on the publication of contentious content, by dissuading users from publishing their tweet deemed offensive by the algorithm developed for the occasion.

Finally, besides that, thousands of tweets and accounts are deleted and suspended every day, based on the rules laid down by the platforms and which are not controlled by anyone.

In other words, Twitter like Facebook have sovereignty over the regulation of the content that is published on their platform, and can therefore do what they want.

It is not possible to leave it to the platforms to tell the truth and the false, to decide what to say and what not to say.

Mathieu slama

Many people, including recognized specialists in these subjects, justify these regulatory efforts by the fact that Twitter like Facebook are private companies, and that they are free to decide what to say or not to say about them. For some, the issue of censorship on social networks is therefore reduced to a simple matter of CGU (the general conditions of use to which one adheres by registering on Twitter or Facebook). But this is not understanding (or worse, pretending not to understand) that these platforms are today hegemonic and that they play a primordial role in contemporary political and democratic debate. In other words, the question of the regulation of content on these platforms has immense democratic implications,and can in no way be reduced to a question of T & Cs.

It is therefore necessary to lay down some clear principles here.

First, that freedom is always preferable to censorship, and therefore that insufficient regulation will always be preferable to excessive regulation.

This principle is fundamental, because it derives from our very conception of freedom.

Then, that it is not possible to leave the platforms to tell the truth and the false, to decide what can be said and what cannot be said, and that any action vis-à-vis of so-called “hateful” content must be taken on the basis of clear regulations enacted by the national legislator.

It is a question of determining very precisely the contours of what is meant by “hateful content”, and adopting the most permissive conception possible.

In other words, it is necessary to limit the restrictions as much as possible, even if it means accepting certain excesses.

Digital capitalism has never been so intrusive, disciplinary and liberticidal.

Mathieu slama

Finally, that the fight against online extremism can in no way turn into a hunt for every syllable used by Internet users, and even less an appeal to denunciation as Facebook is currently experiencing in the United States. How not to worry about the abuses of such a system, where everyone could denounce anyone on any basis, and where, even more seriously, everyone would now pay attention to the slightest word he uses to avoid to be denounced? Because this is the main stake of this affair: that the initiatives taken to fight against hate speech lead us towards a society of permanent suspicion, self-censorship and generalized surveillance of one another.

Digital capitalism has never been so intrusive, disciplinary and liberticidal. It is up to the States not to encourage these abuses (as they are currently doing), but on the contrary to force the GAFA to respect the essential foundations of our democracy, foremost among which is freedom.