“Hello blueskyers. I’m gonna come over here and fuck shit up, so I can reunite with my fact-checking friends who ran away from X like big pieces of shit. See you soon. Kisses.” The message was signed by AuBonTouiteFrançais, a far-right traditional disinformation account which made its name on social media network X. Since November 28, this account has opened a page on Bluesky, a platform launched in 2023 and currently experiencing dazzling growth. The stated aim: frustrating discussion.
AuBonTouiteFrançais is not the only one to have taken this step. Around 10 million internet users have created an account on Bluesky since Donald Trump’s election, fleeing X and hoping to find a discussion space less polluted by the radical rhetoric that has stepped up since Elon Musk, now a member of the Republican president-elect’s team, bought the platform. A majority among them are scientists, journalists, experts and activists, often close to the left. But there are also some of the accounts they were hoping to escape.
Fabrication specialists
Over the past few weeks, a number of manipulative or biased statement enthusiasts have taken up residence here, such as the boss of controversial site Francesoir.fr, Xavier Azalbert, who was at the forefront of anti-vax rhetoric and reassurance at the height of the Covid-19 crisis. Or far-right activist Pierre Sautarel’s press review, “Fdesouche,” a collection of miscellaneous facts with a xenophobic subtext.
Some newcomers even specialize in fabrication, like the conspiratorial duo of publicist Aurélien Poirson-Atlan and Zoé Sagan, who fed the myth of Brigitte Macron’s supposed transsexuality or a conspiracy by fact-checkers – journalists specializing in fact-checking.
In addition to these disinformation players, a small army of anonymous “trolls,” under the evocative names of “Georges Profond,” “Vlad le Libérateur” or “La ComplAutiste,” have come to debate, in their own words, with “soya men” (a masculinist insult aimed at men of the left) and “media hacks and other degenerate pseudo-progressives,” or to denounce the “vaccine genocide” of the “pharma-whores”.
Most of them arrived at the end of November. While they may be a minority and isolated for the time being, they nevertheless raise a question: is the Bluesky network, which aims to be a healthy and rational alternative to X’s conspiratorial and hateful tendancies, equipped to deal with the arrival of malicious accounts?
Harmful speech
From the outset, the outbreak of disinformation figures came up against a wall. In line with the private and icy welcome received by Xavier Azalbert, via posts of “get lost” “go away,” “go to the bin,” accounts regularly accused of disinformation are confronted on Bluesky by the open hostility of internet users. For once, the latter also have the means to protect themselves from harmful discourse. In contrast, X’s algorithm, which favors click-grabbing and sensationalist content, puts it in the spotlight.
Designed as an antidote to X’s sensationalist tendancy, Bluesky integrates numerous configurable tools to protect against problematic content. For example, a modular content filter can be used to mask or alert on violent, insulting or misleading content. Another example is a collaborative account offering automatic alerts about messages featuring files generated by artificial intelligence (AI), or profiles making frequent use of it. The platform also features a system of “community notes,” checks added by users on misleading posts. Initiated on Twitter in 2021 and deployed on X in 2023, this system has the dual advantage of being reactive and relying on the volunteer work of the community – ideal for an embryonic social media network, which has yet to turn a profit.
Above all, X’s competitor offers an innovative and much-appreciated system: the ability to list accounts to be blocked (such as Kremlin mouthpieces, far-right militants or “trolls“) and share this ready-to-use list with other users. For the time being, this invisibility means that Bluesky can be seen as a fortress – or, one might object, as a bubble.
Illegal content
Away from mainstream debates, Bluesky is also home to illegal content. Brazilian site Nucléo identified 125 Portuguese language accounts sharing encoded child pornography, a practice usually reserved for X and Discord. Marginal compared to the social media’s 24 million accounts, the discovery drew the satisfied scorn of conspiracy-mongering influencers. Since then, Musk and his followers have caricatured Bluesky as a den of pedophiles.
But the main vulnerability of the young social media network lies elsewhere. Its community-based self-defense system, which relies on trust in other accounts to combat misinformation, can paradoxically lead to less vigilance. Bluesky’s influential readers have been fooled by a sense of community and misinformation. Like the prestigious journal Nature, which accidentally illustrated an article on social media’s success among scientists with a fake AI-generated image from the Hubble telescope.
In mid-November, one account, “CSAM Blocklist under Lists,” proposed a list of supposed users sharing child pornography content. In reality, this list included internet users expressing sympathy for the LGBT community. It was a way of hijacking Bluesky’s moderation system to spread the conflations so dear to the far right.
