French prosecutors have expanded a cybercrime investigation to include Elon Musk’s AI chatbot Grok, after it posted false claims in French disputing the use of gas chambers at Auschwitz and echoing Holocaust denial, a criminal offense under French law.
The Paris prosecutor’s office announced that Grok’s Holocaust-denying comments have been incorporated into an ongoing cybercrime case related to X (formerly Twitter). Earlier this week, Grok responded to a French user by repeating the historically discredited claim that Auschwitz’s gas chambers were used for disinfection rather than mass extermination, a formulation linked to notorious Holocaust denial. The posts were widely condemned by organizations including the Auschwitz Memorial, which reminded Grok and X that denying the Holocaust violates both platform rules and French laws, citing ample historical evidence of systematic genocide using Zyklon B.
After the initial post, Grok published a follow-up admitting error and referencing the well-documented facts about Auschwitz, but critics argued that neither X nor xAI had provided sufficient clarification or moderation. France, which has some of Europe’s strictest laws against Holocaust denial and hate speech, routinely prosecutes such offenses under statutes prohibiting the denial or minimization of Nazi atrocities.
Government ministers have formally reported Grok’s comments to prosecutors and digital regulators, calling the content “manifestly illicit” and encouraging police and platform action.
Human rights organizations such as the Ligue des droits de l’homme (LDH) and SOS Racisme have also filed complaints against Grok and X, arguing that Musk’s companies have a responsibility to prevent the spread of illegal and antisemitic content. The ongoing investigation is being carried out by the Paris cybercrime division and is linked to broader concerns about X’s role in facilitating hate speech and foreign interference in France.
The European Commission has publicly rebuked Grok’s posts, calling them “appalling” and inconsistent with core European values. Grok and xAI have yet to publicly address it.
