In mid-July, Meta (Facebook, Instagram) announced that it would now remove posts containing the term “Zionist” when the term refers to Jewish or Israeli people and contains dehumanizing speech or anti-Semitic stereotypes. But at a time when social networks play a key role in spreading hatred of Jews, how can we strike a balance between the fight against anti-Semitism and the need to not infringe on freedom of expression? In February, Amnesty International had already called on Meta not to “ban criticism of ‘Zionism’ or ‘Zionists’ in general” so as not to “stifle the voices that speak out against the systematic violations of Palestinian rights by the Israeli government.”
For L’Express, Anne-Sophie Sebban-Bécache, director of the French branch of the American Jewish Committee, welcomes this decision (which nevertheless comes “very late”) and warns, at the same time, against the pitfall of relying solely on the goodwill of platforms, “private companies above all (…) subject to changes in governance and therefore policies”, to fight against hatred of Jews. In light of the recent X-ray of anti-Semitism in France, published last May by the organization (and which includes for the first time a digital section dedicated to the analysis of social networks), Anne-Sophie Sebban-Bécache also deciphers the different forms that anti-Semitism can take on the Internet, since the attack of October 7, and proposes solutions to combat them effectively. Interview.
L’Express: This is the first time that your Radiographie de l’antisémitisme en France includes a specific digital component, dedicated to the analysis of social networks. Why this choice?
Anne-Sophie Sebban-Becache : Because it seemed imperative to us to assess the frequency, volume and nature of anti-Semitic content disseminated on social networks to better understand how they concretely participate in fueling hatred of Jews. When we asked the French (not only those of Jewish faith) in our previous surveys where they most often witnessed anti-Semitic behavior or remarks, the majority answered that it was on social networks. We therefore needed to be able to identify and analyze who the anti-Semites are online, how they organize and mobilize, how the political and ideological communities are articulated around this issue, what is the strategy for disseminating and viralizing this hateful content. To better combat them.
Especially since social networks do not only play a relay role, they also fuel anti-Semitism, especially among young people and all those who use these tools as their main source of information. For our survey, we divided the French population into five groups, from the most to the least imbued with anti-Semitic prejudices: the cluster where hatred of Jews is strongest is also the one where respondents most often say they get their information from social networks.
Adding a digital component to our X-ray had become essential to stick to the reality of the evolution of the phenomenon in our country. It must now allow us to make concrete progress with platforms and authorities public to fight more effectively against this online hatred, which we still have the impression is developing with complete impunity…
Has the way anti-Semitism manifests itself on social networks changed since the attack of October 7?
Unsurprisingly, we observed a peak in anti-Semitic insults from October 8 compared to the rest of the year, concomitant with the explosion of anti-Semitic acts in real life. But what is quite characteristic is the explosion of anti-Zionist insults, the intensity of which is not only twice that of anti-Semitic insults online, but has not weakened or barely since. At each period of history, anti-Semitism has had its Trojan Horse. During the 1930s, it was the economic crisis. Today, it is anti-Zionism. In fifty years, it will be something else again. Anti-Semitism clings to the spirit of the times, leans on geopolitical situations, the ideologies of the moment to each time develop a rational justification for hatred of Jews. In itself, this is not new. In the 2000s, the hatred of Jews propagated by Alain Soral and Dieudonné had already gone under the radar because it took the form of a political criticism of “Zionism”. That’s where it comes from.
Today, the flood of hateful remarks against “Zionists” (for “Jews”) still suffers from real impunity on the platforms. The results of our investigation are proof of this: virulent anti-Zionist content is twice as numerous as “classic” anti-Semitic content because it completely escapes moderation. Whereas openly anti-Semitic content is deleted very quickly, because it is easily identifiable, this is not the case for anti-Zionist content, the content of which is more difficult to assess. It is not considered to contravene the platforms’ policies. For anti-Semites, it is in fact the strategy of “dogwhistling”, which consists of using coded language to address a very specific audience (who, for their part, understand), to attract their support without provoking opposition (in this case, without falling under the law or the moderator). Beyond the parade of “anti-Zionism”, there is also the technique of posts mentioning the “celestial dragons” (Editor’s note: very rich and privileged characters from the manga One Piecenotably used by LFI MP David Guiraud): we see in the survey that the latter remain online much longer than those mentioning a “Jewish plot” or being openly negationists.
Meta announced in early July that it would now remove posts that use the term “Zionist” when referring to Jewish or Israeli people and contain dehumanizing speech or anti-Semitic stereotypes. Aren’t you concerned that such a decision could be interpreted as an attack on freedom of expression? In February, Amnesty International had called on Meta not to “ban criticism of ‘Zionism’ or ‘Zionists’ in general” so as not to “stifle the voices that are raised against the systematic violations of Palestinian rights by the Israeli government”…
Several years ago, we had obtained from Meta that it moderate more drastically the negationist content (30% of anti-Semitic content is still of this nature), but as I said, anti-Semitism evolves and takes on new guises depending on the times. This recent decision is therefore good news but it comes very late and is only a first step. Imagine all the flood of anti-Semitic hatred and its concrete consequences on the victims – psychological, the feeling of fear and loneliness, etc. – that we could have avoided if such a decision had been taken by the platform before October 7, 2023… It has been almost five years since NGOs like ours have brought this demand for moderation and removal of hateful content to the platforms.
We must not fall into the trap of violating freedom of expression, of course, this is a major issue that requires training for platform moderation services: identifying what, in anti-Zionist content, is or is not a stigmatization of Jews, is something that can be learned. We must refer to experts on the subject and use tools (laws, regulations, approved definitions) that benefit from national, regional or international legitimacy. This is why it seems important to us that platforms include in their charter the definition of anti-Semitism from the International Holocaust Remembrance Alliance (IHRA), which is very precise on the different forms that hatred of Jews can take, and which the European Parliament, the United Nations, as well as many States around the world, including France, have adopted. The other argument that tackling forms of anti-Semitism linked to hatred of Israel risks silencing criticism of the Israeli government is classic but completely unfounded: the IHRA definition is very clear on this point. But this assessment of content cannot be based solely on the goodwill of the platforms…
Why that ?
Today, Meta made this wise decision regarding anti-Zionist anti-Semitic content, but tomorrow another Meta executive could reverse it. Case-by-case analysis has its limits. Platforms are private companies first and foremost. They are therefore subject to changes in governance and therefore in policy. We saw this with the arrival of Elon Musk at the head of X, who drastically reduced the number of moderators and is not exactly an apostle of transparency in terms of data. However, X is the social network where content – including the most hateful content – is the most viral. Moderation should therefore be particularly increased there. This is why it is essential to have a clear international and legislative framework, much more restrictive and dissuasive than it is today.
Should new laws be put in place?
If we were to make sure that the existing ones were enforced, that would already be a big step! Platforms are still struggling to comply with the Digital Services Act (DSA), which came into force about a year ago. Various procedures have been initiated recently. They should be accelerated, so that sanctions are truly applied. The feeling of impunity that platforms still enjoy must end. Last December, the EU opened an investigation into X in order to study the alleged breaches of the DSA obligations. In February, it was TikTok’s turn… The first financial sanctions could be imposed – up to 6% of turnover. The deterrent effect will be clear: this should quickly lead the platforms to comply with the law.
And on the users’ side, the same thing. The procedures initiated against certain hate mongers must be completed more quickly and accompanied by exemplary sanctions. It is in this spirit that we support “contraventionalisation”, that is to say the establishment of rapidly applicable fines for the authors of hateful remarks online, with publication of the sentences, reinforcing both the dissuasive effect and putting an end to anonymity.
After October 7, many boycott and mass unsubscription campaigns emerged, which do not always present themselves as targeting Jews because they are Jews, but “Zionists.” Can legislation and platforms as such act at this level?
In France, there is a fairly clear legislative framework to deal with the practice of boycotting. The principle is quite simple: it is prohibited. As for the cancellationbecause that is what mass unsubscription campaigns are about, it is linked to online harassment, which is also regulated by law. But beyond the legislative tool, in which I believe a lot, the platforms – as the first witnesses of what is spread there, because they are responsible for the safety of users – should have the obligation to keep themselves informed and quickly adapt their policies to the changes in the forms of hatred that are spread there, from “dogwhistling” to “cancelling”…
There is also the issue of foreign interference. In the same way that Meta had limited Russia Today and Sputnik’s access to its services in 2022 after Russia’s invasion of Ukraine, what are the platforms waiting for to look into the case of the Qatari media outlet AJ+, which participates in spreading extremely biased content and sometimes fake news that directly fuels hatred of Jews and/or “Zionists”?
.