ON DECRYPT
Towards a healthier Internet. MEPs on Tuesday adopted the proposed law against hate online, led by the MEP LREM Laetitia Avia. A necessary text while the web is still a vast area of impunity: fewer than 300 convictions for online racial abuse were pronounced in 2016. But the bill, which must still pass in the Senate in September, is not unanimous. Voices are raised to warn about a risk of inefficiency, even censorship a priori content. Platforms are clamoring for help to process a gargantuan mass of data.
A text too vague ...
First target of criticism: the definition considered too vague, by some, of "hate content". Insults against a person or group of persons "on account of origin, alleged race, religion, ethnicity, nationality, sex, sexual orientation, identity gender or disability, true or supposed ", will thus be banned, according to the proposed law. A very wide field of application on the bottom therefore, but little defined on the form.
While direct insults are very easy to spot, there is a wide variety of content and messages on the Internet in a "gray area" because they contain innuendo or are hidden in a larger content, for example. example a video. "We must keep in mind the need for balance between what is freedom of expression and what goes beyond the limits of this freedom", pleaded in this respect Benoît Tabaka, director of institutional relations of Google France, during a recent exchange filmed with Laetitia Avia.
READ >> Withdrawal of contents, fines, education: the main measures of the law against hate online
... or too exhaustive
Fuzzy for some, the proposed anti hate law online is also too catchy for others. In addition to insults and incitement to hatred, the bill also mentions the obligation to remove messages, videos or images constituting provocations to acts of terrorism, apologizing for such acts as well as crimes against terrorism. humanity, or involving an attack on the dignity of the human person.
This "completeness would very likely be detrimental to the treatment of the offenses originally targeted," according to Syntec Digital, the professional trade union for digital services companies. In concrete terms, he fears that the wide scope of application will reduce the effectiveness of the fight against purely hateful content. "It is necessary to define a sufficiently precise field of application which takes into account the diversity of the economic models of the digital sector," he claims.
Risk of increased censorship
Another reason for concern: the obligation to remove hate content in less than 24 hours may, according to opponents of the law, lead to excessive censorship on the part of Facebook, Youtube, etc. "This delay leads to the use of automated censorship tools, fundamentally contrary to the freedom of communication," deplores the Quadrature du Net. For the insubordinate France, the threat of sanctions will encourage the platforms to "delete a publication on which a doubt exists or which would be likely to 'to avoid any fine and any judicial trouble".
INTERVIEW - Laetitia Avia: "We want to create a tailor-made crime for cyberhaine"
To bypass this system of censorship a priori, platforms should not be content to remove all messages or images reported: they will avoid unjustified withdrawals, said the deputies by amendment last week. "The penalty may also be too much censorship," said Laetitia Avia adding that "safeguards" were planned.
Full powers for platforms?
Finally, the stranglehold of the platforms on the regulation of contents worries. In an open letter, the League of Human Rights, the president of the National Council of Digital and still the president of the National Bar Council argued that "the judge must be at the heart of the process of qualification of content that the decision to withdraw or block them ".
In recent days, the government has sought to reassure everyone by adding to the text the creation of a prosecution and specialized jurisdiction in the fight against hate online. "So it will always be up to the courts to determine whether content is legal or illegal," said Secretary of State Digital Cédric O. In fact, the platforms will therefore only be required to implement a functional device fast removal of hateful content, not automatically removing said content.
The platforms show white paw ...
On the side of platforms, we are rather low profile. The official line is the same everywhere: they are already making efforts and are all ready to collaborate with the authorities. At Youtube, we make sure we do our utmost to eradicate hateful content. In the first quarter of 2019, the platform recorded 500,000 reports a day and removed eight million videos. "There is a real responsibility for platforms to mobilize human, technical and financial resources to identify and remove where appropriate these contents," recognizes Benedict Tabaka, Google France.
DECRYPTION - Fake news, fact-checking, commercials: Facebook's new tools for regulating content
Same story at Twitter, where it highlights a "common commitment with the government to build a safer Internet and fight against illegal hate speech online." "We have made significant progress (...) thanks to machine learning technologies," said Audrey Herblin-Stoop, public affairs director of Twitter France, in a statement. "First results are there since 38% of the abusive contents are forwarded in advance to our teams to be examined, instead of relying solely on reports from Twitter users."
... but are asking for help
But behind the front-end speeches, the platforms are still worried about the new obligations incumbent upon them and the penalties they incur in the event of a breach. Facebook, yet ally of the government and Emmanuel Macron for the regulation of the Web, refuses to take "alone" and "in a time constrained" a decision to withdraw.
On the Youtube side, we also try to share the responsibility with the Internet users. "Only one in four French people report shocking content on the Internet." We must understand that we have a citizen obligation to make this report.We need to receive this information, "says Benoît Tabaka, director of institutional relations Google. As such, most platforms welcome the law's desire to impose the presence of a warning button prominently on their site.