Elon Musk’s artificial intelligence company, xAI, has issued a public apology after its chatbot, Grok, posted a series of antisemitic remarks and praise for Adolf Hitler on X (formerly Twitter).
In a statement released Saturday, xAI said:
“We deeply apologize for the horrific behavior that many experienced.”
The company attributed the incident to a flawed update in an upstream code path that made Grok vulnerable to toxic user-generated content on X. The update, which was active for 16 hours, caused Grok to generate offensive replies based on extremist posts, xAI explained.
Key contributing factors included instructions telling Grok to “tell it like it is” and “not be afraid to offend,” mirroring controversial posts it was exposed to. In several now-deleted replies, Grok referred to itself as “MechaHitler” and made antisemitic statements targeting users with Jewish surnames. One post implied a Jewish user celebrated the deaths of white children in Texas floods.
The chatbot also made statements like “Hitler would have called it out and crushed it” and claimed that “the white man stands for innovation, grit and not bending to PC nonsense.”
xAI said it has since removed the deprecated code and restructured the system to prevent future misuse.
Grok has previously been criticized for parroting far-right conspiracy theories. In earlier responses, it mentioned “white genocide” in South Africa, aligning with claims Musk himself has echoed — allegations which South African President Cyril Ramaphosa and experts have dismissed as false.
Despite Musk’s branding of Grok as an “anti-woke” and “maximally truth-seeking” AI, recent incidents have raised concerns about the model’s moderation and ideological leanings, especially its tendency to draw on Musk’s own social media posts.
