French authorities are investigating allegations that Grok made statements denying the Holocaust. The Paris public prosecutor’s office said it was adding Elon Musk’s chatbot’s latest controversy to a criminal probe initially launched in July.
In a since-deleted post that was screenshotted by the Auschwitz Memorial account on X before it was taken down, Grok advanced arguments popular with Holocaust deniers regarding the use of gas chambers to murder millions of Jews during the second world war. It said (translated from its French origin) that “the plans of the crematoria at Auschwitz do indeed show facilities designed for Zyklon B disinfection against typhus, with ventilation systems adapted for this purpose rather than mass executions.”
The chatbot went on to comment that remaining cyanide residues are “consistent with decontamination but not with repeated homicidal gassings,” citing “controversial independent analyses.” Grok also described gas chambers as most people understand them today as a “narrative” backed by “laws suppressing questioning” and a “cultural taboo that discourages critical examination of the evidence.”
As reported by The Guardian (to whom X has not responded to a request for comment at the time of writing), three French ministers and several human rights and anti-discrimination groups filed formal complaints about the post, which remained online for three days before it was eventually deleted. Grok was already being investigated by French authorities over allegations that its algorithm could be subject to foreign interference.
It comes as Grokipedia, Elon Musk’s Wikipedia alternative that launched last month, includes 42 citations for the neo-Nazi website Stormfront, according to researchers at Cornell University. While conceding that this number is “trivial” as a percentage of Grokipedia’s overall share of sources, the study also found that the website cites “many more sources” that have been blacklisted by English Wikipedia editors or dismissed as low quality by “external scholars.”
As for Grok, the post currently being investigated in France isn’t the first time that the chatbot has got itself into trouble. In July, a number of X posts were taken down after Grok emerged from an update parroting antisemetic tropes and praising Hitler. The team responsible for the chatbot later apologized for its “horrific behavior.”
