France will investigate Musk’s Grok chatbot after Holocaust denial claims

PARIS (AP) — France’s government is taking action against billionaire Elon Musk ‘s artificial intelligence chatbot Grok after it generated French-language posts that questioned the use of gas chambers at Auschwitz, officials said.

Grok, built by Musk’s company xAI and integrated into his social media platform X, wrote in a widely shared post 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 long associated with Holocaust denial.

The Auschwitz Memorial highlighted the exchange on X, saying that the response distorted historical fact and violated the platform’s rules.

In later posts on its X account, the chatbot acknowledged that its earlier reply to an X user was wrong, said it had been deleted and pointed to historical evidence that Auschwitz’s gas chambers using Zyklon B were used to murder more than 1 million people. The follow-ups were not accompanied by any clarification from X.

In tests run by The Associated Press on Friday, its responses to questions about Auschwitz appeared to give historically accurate information.

Grok has a history of making antisemitic comments. Earlier this year, Musk’s company took down posts from the chatbot that appeared to praise Adolf Hitler after complaints about antisemitic content.

The Paris prosecutor’s office confirmed to The Associated Press on Friday that the Holocaust-denial comments have been added to an existing cybercrime investigation into X. The case was opened earlier this year after French officials raised concerns that the platform’s algorithm could be used for foreign interference.

Prosecutors said that Grok’s remarks are now part of the investigation, and that “the functioning of the AI will be examined.”

France has one of Europe’s toughest Holocaust denial laws. Contesting the reality or genocidal nature of Nazi crimes can be prosecuted as a crime, alongside other forms of incitement to racial hatred.

Several French ministers, including Industry Minister Roland Lescure, have also reported Grok’s posts to the Paris prosecutor under a provision that requires public officials to flag possible crimes. In a government statement, they described the AI-generated content as “manifestly illicit,” saying it could amount to racially motivated defamation and the denial of crimes against humanity.

French authorities referred the posts to a national police platform for illegal online content and alerted France’s digital regulator over suspected breaches of the European Union’s Digital Services Act.

The case adds to pressure from Brussels. This week, the European Commission, the EU’s executive branch, said that the bloc is in contact with X about Grok and called some of the chatbot’s output “appalling,” saying it runs against Europe’s fundamental rights and values.

Two French rights groups, the Ligue des droits de l’Homme and SOS Racisme, have filed a criminal complaint accusing Grok and X of contesting crimes against humanity.

X and its AI unit, xAI, did not immediately respond to requests for comment.