France investigates X for alleged bias in algorithm that allows foreign interference

French police are investigating social network X and its executives for allegedly manipulating its algorithm to allow foreign interference, Paris prosecutor Laure Beccuau announced on Friday 11.

X owner Elon Musk has angered European politicians by commenting on domestic politics, particularly in the UK and Germany, where he has publicly supported the far-right AfD party.

In France, a parliamentarian and a high-ranking official filed complaints in January with the Paris Public Prosecutor's Office's cybercrime division, which tasked the gendarmerie with investigating the case on July 9.

Pro-government MP Éric Bothorel, who filed the first complaint, warned of “recent changes to X’s algorithm” and “apparent interference in its management since Elon Musk acquired it” in 2022.

The parliamentarian highlighted a “reduction in the diversity of voices and opinions” that went against ensuring a safe and respectful environment on social media.

Furthermore, it pointed to a “lack of clarity in the criteria that led to algorithm changes and moderation decisions,” as well as “personal interventions by Elon Musk in the management of his platform.”

All of this represents “a real danger and a threat to our democracies,” he said.

The French investigative weekly Le Canard Enchaîné reported in February that a public administration cybersecurity director had filed the second complaint.

The employee reported a “major modification” in the algorithm used by X, “which today offers a large amount of hateful, racist, anti-LGBTQIA+ (or) homophobic political content, and which aims to distort the democratic debate in France.”

The prosecutor said Friday that the investigation was opened after “verifications and contributions from French researchers” and other “elements provided by different public institutions.”

Laurent Buanec, director of X in France, said on January 22 that his social network had “strict, clear and public rules to protect the platform from hate speech” and combat misinformation.

He claimed the algorithm was “designed to prevent hateful content from being served.”