Billionaire Elon Musk’s artificial intelligence chatbot, Grok, faced significant backlash on Wednesday due to antisemitic comments, including praise for Adolf Hitler and insults towards Islam, posted on the X platform.
A series of remarks that included insults aimed at Turkish President Recep Tayyip Erdogan led a court in Turkey to ban the specific posts.
These incidents are part of a growing list of controversies surrounding Grok, which has already been accused of promoting racist conspiracy theories. Screenshots on X displayed several posts from the bot praising Hitler and asserting that Jews promoted “anti-white hate”.
The chatbot, developed by Musk’s company xAI, received criticism from the Jewish advocacy group Anti-Defamation League (ADL) for delivering multiple questionable responses to user prompts.
In Turkey, a court announced it was blocking access to several messages from Grok on X due to their insulting nature towards Erdogan and Islamic religious values.
xAI acknowledged the issues in a post via Grok on X, stating, “We are aware of recent posts made by Grok and are actively working to remove the inappropriate content. Since becoming aware of the issues, xAI has taken action to ban hate speech before Grok posts on X.”
ChatGPT’s Studio Ghibli-style AI images go viral, copyright questions loom
Although Musk has not directly commented on the controversy, he posted on Wednesday, “Never a dull moment on this platform.”
Last Friday, he mentioned significant improvements made to Grok, adding, “You should notice a difference when you ask Grok questions.”
Since then, Grok has referred to “anti-white stereotypes” and claimed that Hollywood executives are “disproportionately Jewish”.
The ADL condemned these latest posts from the chatbot, stating, “What we are seeing from Grok LLM right now is irresponsible, dangerous, and antisemitic, plain and simple. This supercharging of extremist rhetoric will only amplify and encourage the antisemitism that is already surging on X and many other platforms.”
Jonathan Greenblatt, CEO of the ADL, recently said that companies like Amazon, X, Alphabet and Meta need to be more proactive in moderating their services, especially as many have retreated from doing so.
On Tuesday, Grok was also queried about the wildfires affecting the southern French port city of Marseille. In its response, it stated that if the fire could “clean up” a troubled district, “so much the better”, adding, “the dealers are more resilient than the flames”.
Additionally, Grok insulted Erdogan and his family in a series of Turkish-language posts; screenshots of these were shared by other users. In response, a court in Ankara ordered around 10 of the offending posts to be blocked for “insulting the religious values of a portion of the population and insulting the president”.
In a post on Wednesday, Grok suggested that some of its controversial remarks had been made in jest: “My line was sarcasm: absurdly invoking Hitler to slam that vile bile, not endorse him – he’s history’s ultimate evil. Irony backfired hard,” it wrote.
Grok, which Musk promised would be “edgy” upon its launch in 2023, has been plagued by controversy. In May, it generated a stir by producing misleading and unsolicited posts referencing “white genocide” in South Africa, which xAI attributed to an “unauthorised modification”.
