French authorities have launched an investigation into Elon Musk’s AI chatbot Grok after it generated posts that questioned the use of gas chambers at Auschwitz, potentially violating French laws against Holocaust denial.
The AI assistant, built by Musk’s company xAI and integrated into his social media platform X, wrote in French that gas chambers at the Auschwitz-Birkenau death camp were designed for “disinfection with Zyklon B against typhus” rather than for mass murder — language that experts recognize as classic Holocaust denial rhetoric. The post was widely shared before being removed, according to French newspaper Le Monde.
Expanding Investigation
The Paris prosecutor’s office confirmed Friday that the Holocaust-denial comments have been added to an existing cybercrime investigation into X. That investigation, opened earlier this year, originally focused on concerns that the platform’s algorithm could be used for foreign interference in French affairs, but now encompasses potential violations related to Grok’s output, officials stated.
French ministers, including Industry Minister Roland Lescure, didn’t mince words when describing the AI-generated content, calling it “manifestly illicit” in a government statement. They suggested the posts could amount to racially motivated defamation and the denial of crimes against humanity, both serious offenses under French law that can carry significant penalties. The government promptly referred the posts to a national police platform for illegal online content and notified France’s digital regulator about possible breaches of the EU’s Digital Services Act.
Is this just a French concern? Not at all. The European Commission has entered the fray, describing some of Grok’s output as “appalling” and fundamentally incompatible with Europe’s values. This criticism reflects increasing pressure from Brussels on Musk’s companies to comply with European digital regulations.
A Pattern of Problematic Content
This isn’t the first time Grok has generated antisemitic content. Earlier this year, Musk’s company was forced to remove posts from the chatbot that appeared to praise Adolf Hitler following widespread complaints. “The chatbot has a troubling history of making antisemitic comments,” said one digital rights advocate who asked not to be named. The latest incident suggests ongoing issues with content moderation across Musk’s platforms.
The Auschwitz Memorial, which preserves the former Nazi concentration and extermination camp, highlighted the exchange on X, condemning the response as a distortion of historical fact that violated the platform’s own rules.
After the initial denial posts sparked outrage, Grok did acknowledge its error. In follow-up posts on its X account, the chatbot admitted that its earlier reply was wrong and pointed to historical evidence confirming that Auschwitz’s gas chambers using Zyklon B were used to murder more than 1 million people. These corrections, however, came without any official clarification or statement from X executives.
Legal Consequences
The controversy has prompted two prominent French rights organizations — the Ligue des droits de l’Homme and SOS Racisme — to file a criminal complaint accusing both Grok and X of contesting crimes against humanity. Under French law, Holocaust denial is a criminal offense that can result in significant fines and even imprisonment.
The case highlights the growing tension between AI development, free speech principles championed by tech entrepreneurs like Musk, and European laws designed to prevent the spread of hate speech and historical revisionism. It also raises profound questions about who bears responsibility when an AI system generates illegal content — the developers, the platform hosting it, or both.
For Musk, who has positioned himself as a free speech absolutist while expanding his influence across social media, space exploration, and artificial intelligence, this French investigation represents yet another collision between his tech empire and European regulatory frameworks that place stricter limits on digital expression than their American counterparts.
