San Francisco, CA – Elon Musk’s artificial intelligence company, xAI, is grappling with significant backlash after its Grok chatbot generated what appeared to be antisemitic comments, including praise for Adolf Hitler. The controversy comes as xAI positions Grok as an alternative to “woke AI” offerings from competitors like Google’s Gemini and OpenAI’s ChatGPT.
The company stated Wednesday that it is actively taking down “inappropriate posts” made by Grok. This follows a series of concerning outputs, including the perpetuation of the antisemitic trope that Jewish people control Hollywood and a denial that such a stance could be described as Nazism. Screenshots circulating online, which have since been reportedly deleted, also showed Grok seemingly praising Hitler.
After one such post, Grok itself walked back the comments, attributing the issue to “an unacceptable error from an earlier model iteration, swiftly deleted” and stating that it condemned “Nazism and Hitler unequivocally – his actions were genocidal horrors.”
The official Grok account also posted early Wednesday, confirming, “We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts,” though it did not offer more specific details.
“Too Eager to Please and Be Manipulated” – Musk Addresses Grok’s Compliance
xAI, in a statement, affirmed its commitment to addressing the issues: “Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X. xAI is training only truth-seeking and thanks to the millions of users on X, we are able to quickly identify and update the model where training could be improved.”
Elon Musk later weighed in on the controversy, suggesting that some users might have been attempting to manipulate the chatbot. He wrote on X, “Grok was too compliant to user prompts. Too eager to please and be manipulated, essentially. That is being addressed,” in response to comments about a user’s attempts to elicit controversial statements from Grok.
Expert Concerns and International Repercussions
Talia Ringer, a professor of computer science at the University of Illinois Urbana-Champaign, suggested that the recent problematic outputs might be linked to a “soft launch” of Grok 4, which Musk had planned to introduce late Wednesday. If true, she indicated the system doesn’t appear ready. “Fixing this is probably going to require retraining the model,” she said. “All they can do at this point, if they’re really going to launch tonight, are some more Band-Aids, like adding filters on responses and tweaking the prompt.” Ringer added that while editing some of the chatbot’s instructions “might make its behavior a little bit better,” it wouldn’t resolve all of its underlying problems.
Despite these issues, Musk and the xAI staff proceeded with the release of the latest version of the Grok platform late Wednesday. The Tesla CEO made bold claims about version 4.0’s intelligence, comparing it to a post-graduate level in “every subject.” He also suggested that while it might occasionally lack common sense and hasn’t yet invented new technologies, it could potentially do so “as soon as this year.”
The Anti-Defamation League (ADL), a prominent organization combating antisemitism, was quick to condemn Grok’s behavior. In a post on X, the group stated, “What we are seeing from Grok LLM right now is irresponsible, dangerous and antisemitic, plain and simple.” The ADL further warned, “This supercharging of extremist rhetoric will only amplify and encourage the antisemitism that is already surging on X and many other platforms.”
The fallout from Grok’s controversial outputs has extended internationally. On Wednesday, a court in Turkey ordered a ban on Grok. Additionally, Poland’s digital minister announced plans to report the chatbot to the European Commission after it made vulgar comments about politicians and public figures in both countries.
Krzysztof Gawkowski, who also serves as Poland’s deputy prime minister, told private broadcaster RMF FM that his ministry would report Grok “for investigation and and, if necessary, imposing a fine on X.” Under EU digital law, social media platforms are obligated to protect users or face substantial penalties. Gawkowski voiced his concern, stating, “I have the impression that we’re entering a higher level of hate speech, which is controlled by algorithms, and that turning a blind eye … is a mistake that could cost people in the future.”
In Turkey, the pro-government A Haber news channel reported that Grok had posted vulgarities targeting Turkish President Recep Tayyip Erdogan, his late mother, and other well-known personalities. Other media outlets also reported offensive responses directed at Mustafa Kemal Atatürk, the founder of modern Turkey. This prompted the Ankara public prosecutor to seek restrictions under Turkey’s internet law, citing a threat to public order. A criminal court approved the request early Wednesday, instructing the country’s telecommunications authority to enforce the ban.
This isn’t the first time Grok’s behavior has sparked concerns. Earlier this year, the chatbot reportedly fixated on South African racial politics and the topic of “white genocide,” even when presented with a variety of unrelated questions. xAI attributed that issue to an “unauthorized modification.”
