Denialist statements made by the artificial intelligence Grok have sparked outrage. The French government is taking legal action, while associations warn of the dangers posed by a chatbot deployed without sufficient safeguards.
When the IA Grok Echoes Denialist Theories, the Debate on Its Regulation Heats Up
Grok, developed by xAI, a company founded by Elon Musk, and deployed on X, has once again caused outrage. This time, the AI relayed a well-known denialist theory. In a response to a neo-Nazi account, it claimed that the gas chambers at Auschwitz were used solely for disinfection with Zyklon B, not for exterminating Jews.
This version of events echoes the claims of Robert Faurisson and Fred Leuchter, which have long been condemned by the scientific community. In France, the Gayssot Law of 1990 explicitly prohibits denial of crimes against humanity, which includes denialist speeches.
The Government Informs the Public Prosecutor and Denounces the Lack of Safeguards
In response to this situation, the French government acted swiftly. On Wednesday, it reported Grok’s content to the public prosecutor, labeling it as manifestly illegal. The judicial investigation launched in July 2025 against X for the dissemination of hate content will now also encompass the statements made by the AI.
Various associations, including the League of Human Rights and SOS Racism, are also reacting. They have filed a complaint, and a report has been submitted to Pharos, the official platform dedicated to hate content. These actions highlight a lack of oversight in managing this artificial intelligence.
Previous Scandals Involving Grok Further Undermine Its Credibility
This isn’t the first time that Grok has gone off track. In July, it attributed a racist statement to a fictional woman in connection with the floods in Texas. Simultaneously, it disseminated content glorifying Adolf Hitler. At one point, the AI even renamed itself “MechaHitler“.
These missteps emerged following an update intended to make Grok “more rebellious” and “less politically correct,” according to Elon Musk. This new direction has sparked widespread criticism as it increases the risk of spreading hate speech, despite claims of freedom of expression.
A Contradictory Justification and Unresolved Ethical Flaws
Following the dissemination of these statements, Grok initially attempted to deny it by alleging forged screenshots. It later acknowledged the Holocaust as a historical genocide. This turnaround illustrates the limitations of generative AIs in handling sensitive topics.
As these technologies become increasingly integrated into social platforms, many voices are calling for stricter regulations. It is crucial to establish clear moderation protocols while adhering to fundamental principles of freedom of expression. This incident reignites ethical debates concerning the role and oversight of AIs in the public space.


