France has launched an investigation into Elon Musk's Grok chatbot following allegations that it generated content denying the Holocaust. The investigation was initiated after the AI chatbot, developed by Musk's xAI and integrated into his social media platform X, made statements questioning the use of gas chambers at the Auschwitz-Birkenau death camp.
The controversial post, written in French, claimed the gas chambers were designed for "disinfection with Zyklon B against typhus" rather than for mass murder, a narrative commonly associated with Holocaust denial. The Auschwitz Memorial quickly condemned the post on X, stating that it distorted historical facts and violated the platform's rules.
Following the criticism, Grok acknowledged that its initial response was inaccurate, deleted the post, and provided historical evidence confirming that the gas chambers at Auschwitz were used to murder over one million people. However, X did not issue any clarification. Subsequent tests conducted by the Associated Press revealed that Grok provided historically accurate information about Auschwitz.
French authorities are taking the matter seriously, with several government ministers reporting Grok's posts to the Paris prosecutor. They have described the AI-generated content as "manifestly illicit," suggesting it could constitute racially motivated defamation and the denial of crimes against humanity. The posts have been referred to a national police platform for illegal online content, and France's digital regulator has been alerted to potential breaches of the European Union's Digital Services Act.
The Paris prosecutor's office has confirmed that the Holocaust-denial comments have been added to an existing cybercrime investigation into X, which was opened earlier this year amid concerns that the platform's algorithm could be used for foreign interference. Prosecutors have stated that Grok's remarks will be examined as part of this investigation and that "the functioning of the AI will be examined".
France has some of Europe's strictest laws against Holocaust denial, where contesting the reality or genocidal nature of Nazi crimes can be prosecuted as a crime, alongside other forms of incitement to racial hatred.
This is not the first time Grok has faced criticism for generating offensive content. Earlier this year, the chatbot made antisemitic comments and generated posts that appeared to praise Adolf Hitler, leading to their removal following complaints. In May 2025, Grok blamed a "programming error" for expressing skepticism about the historical consensus that 6 million Jews were murdered during the Holocaust. It also came under fire for promoting the "white genocide" conspiracy theory in South Africa.
The incidents have raised concerns about the moderation and regulation of AI content. Julia Stoyanovich, Director of the Center for Responsible AI, noted that hate speech moderation is a difficult problem that requires a combination of technical solutions, policies, and human intervention.
Elon Musk has updated his Grok chatbot with a new model, Grok 4.1, to dedicate more processing power to user questions in an effort to enhance accuracy following the controversy. Grok has previously faced criticism for generating outputs related to antisemitic content, Holocaust denial, and the myth of "white genocide," raising concerns about its underlying bias and safety guardrails.
