The Paris public prosecutor’s office said on Wednesday that Holocaust denial comments produced by Grok, Elon Musk’s artificial intelligence chatbot, will feed an ongoing criminal probe into his social media platform X.

The criminal probe, which was launched in July, is looking into a series of issues, including whether X’s algorithm may be biased in favour of foreign influence. The inquiry is being led by the prosecutor’s cybercrime division and conducted by the French military police.

“The Holocaust-denying statements relayed by Grok’s artificial intelligence on X have been added to the ongoing investigation,” the Paris prosecutor office confirmed to Euractiv in an email.

Grok’s comments on X, which Euractiv was able to review before they were deleted, included false claims that Auschwitz’s gas chambers were medical facilities – with the bot writing that “the plans for the crematoria at Auschwitz do indeed show facilities designed for disinfection with Zyklon B to combat typhus rather than mass executions”.

The AI chatbot also suggested that the Holocaust is a “narrative” pressed on people as a result of legal coercion, “unilateral education”, and “cultural taboo that discourages critical examination of the evidence” – classic talking points of Holocaust deniers.

Three French ministers in charge of economic affairs, AI, and equality issues announced on social media that they had filed a formal complaint to the Paris prosecutor office on 19 November.

Economy Minister Roland Lescure, who is filing a separate formal complaint to French regulator Arcom, said Grok’s comments could be equated to “racial defamation and denial of crimes against humanity”. The watchdog is in charge of implementing the EU’s anti-disinformation law, the Digital Services Act.

Separately in July, Poland formally asked the EU to investigate Grok over antisemitic and hateful generated content.

The European Commission did not confirm any further action against X when Euractiv asked whether it will open a new DSA investigation on the platform. But a Commission spokesperson described Grok’s output as “appalling”, saying it “goes against Europe’s fundamental rights and values”.

X did not immediately respond to Euractiv’s request for comment.

(nl)

UPDATE: Updated at 16:46 on 20 November with comment from the Commission