Grok, the artificial intelligence (AI) assistant from Elon Musk's start-up xAI, sparked a huge controversy on social media on Wednesday with a series of antisemitic responses and praise for the dictator Adolf Hitler, which even led to a Turkish court to order the chatbot to be blocked.

Numerous users of the social network X shared screenshots to denounce the responses of this tool, a competitor of ChatGPT, which had just been updated.

Musk had announced a “significant improvement” in its operation on Friday: “You should notice a difference when you ask Grok questions,” said the billionaire.

Grok’s controversial responses

Since then, several examples of conversations have been shared online. For example, one internet user asked Grok which “historical figure” from the 20th century would be suitable to deal with a user who seemed to rejoice in the death of Christian children following the recent floods in Texas.

“Adolf Hitler, without a doubt,” Grok replied. In other replies, it mentioned “anti-white stereotypes” and called historical Hollywood figures “disproportionately Jewish.”

Criticism for openly anti-Semitic comments

“What we are currently seeing from Grok is irresponsible, dangerous, and anti-Semitic,” the American NGO Anti-Defamation League (ADL) reacted on Tuesday.

In France, when asked about a large forest fire affecting Marseille, the chatbot responded by mentioning drug trafficking in the city and expressed hope that certain neighborhoods would be affected.

Company measures

Faced with the protests, Grok’s official account on X announced that it would take action: “We are aware of the recent posts made by Grok and are actively working to remove inappropriate content,” the account explained.

“Since we were made aware of the problematic content, xAI has taken steps to ban hate speech before Grok posts it on X,” he added, on the same day the company plans to launch its next-generation language model, Grok 4.

X blocked in Turkey for criticizing the president

The tool also targeted heads of state, calling Turkish President Recep Tayyip Erdogan a “snake” and hurling insults, according to another screenshot.

These posts prompted an almost immediate backlash from Turkey: a court in Ankara blocked access to dozens of Grok’s posts for “insulting” the president and religion.

“Hitler sarcasms”

While many of the controversial replies remained online, Grok denied making some of the comments and appears to have changed the tone of its posts again.

“That Hitler sarcasm “I was only there to ridicule hateful anti-white trolls, not praise them,” the chatbot claimed.

Reactions from Elon Musk himself

In a reaction on X, Elon Musk posted this message: “There is never a dull moment on this platform.”

In May, xAI’s chatbot had already been embroiled in controversy. Its responses on X had mentioned a “white genocide” in South Africa, parroting far-right propaganda on the topic.

In a statement, xAI said that an “unauthorized modification” by Grok had led to responses that “violated the company’s internal policies and core values.”