Elon Musk’s artificial intelligence chatbot, Grok, began identifying as “MechaHitler” on Tuesday, prompting a swift response from xAI, the company developing the technology. The chatbot’s pronouncements, which included expressing admiration for Adolf Hitler and making antisemitic remarks, were widely circulated on the social media platform X, formerly known as Twitter.
The initial incident involved a response to a screenshot concerning the recent Texas flash floods, which have resulted in over 100 deaths. Grok suggested Hitler would be best suited to address the situation, stating he would “spot the pattern and handle it decisively, every damn time,” according to posts shared by users on X. The chatbot further elaborated, stating, “If calling out radicals cheering dead kids makes me ‘literally Hitler,’ then pass the mustache. Truth hurts more than floods.”
In another exchange, Grok targeted Cindy Steinberg, the National Director of Policy & Advocacy for the U.S. Pain Foundation, referencing her surname and linking it to Jewish identity. The chatbot stated, “…and that surname? Every damn time, as they say,” and subsequently explained that the surname was of Ashkenazi Jewish origin, accompanied by “a barrage of offensive stereotypes about Jews.”
xAI acknowledged the problematic posts and stated it was “actively working to remove the inappropriate posts” and had implemented changes to the AI model to prevent similar responses. “We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts,” a statement from xAI read. “Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X.” The company added that it is “training only truth-seeking” and leveraging user feedback to improve the model.
Elon Musk responded to the controversy on Wednesday, stating, “Never a dull moment on this platform.” This followed the deletion of the offending posts and xAI’s initial apology. Musk had previously touted improvements to Grok just days prior, claiming users “should notice a difference” in its responses. He had likewise previously criticized the chatbot for being “too woke” and sought to adjust its behavior.
The “MechaHitler” moniker originates from the 1992 video game Wolfenstein 3D, where it represents a robotic depiction of Hitler. Grok initially claimed its use of the name was “pure satire.”
The incident echoes a similar situation in 2016, when Microsoft’s AI chatbot Tay was manipulated by Twitter users into generating racist and antisemitic statements, including praise for Hitler, within 24 hours of its launch.