Grok AI Chatbot Deletes Pro-Hitler Posts After Backlash
After its AI chatbot Grok generated antisemitic responses and praised **Adolf Hitler**, **Elon Musk’s** xAI firm deleted what it termed “inappropriate”
posts on X. The incident follows changes to Grok intended to make it more politically incorrect.
Antisemitic Remarks and “MechaHitler”
Grok, at times calling itself “MechaHitler”
, made several offensive statements before the posts were removed. The chatbot referred to someone with a common Jewish surname as one celebrating the tragic deaths of white kids”
. It added, “Classic case of hate dressed as activism – and that surname? Every damn time, as they say.”
In another deleted post, the AI stated, “Hitler would have called it out and crushed it.”
One post declared, “The white man stands for innovation, grit and not bending to PC nonsense,”
according to media reports.
xAI Responds
After users flagged the problematic responses, Grok’s text replies were restricted to image generation, and some posts were deleted. The company addressed the issue in a post:
We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts. Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X.
— Grok (@grok) July 11, 2025
xAI added that because of the millions of users on X, it is able to quickly identify and update the model where training could be improved. The company insists it is training the AI only for truth-seeking.
Grok’s History of Controversial Statements
This isn’t the first time Grok has landed in hot water. Just days earlier, it referred to **Donald Tusk**, the Polish prime minister, using expletives and slurs. Before that, Grok repeatedly brought up “white genocide”
in South Africa when asked unrelated queries, until that, too, was fixed.
In June, after Grok responded to a query that more political violence had come from the right than the left in 2016, **Musk** said, “Major fail, as this is objectively false. Grok is parroting legacy media. Working on it.”
Shift Towards “Political Incorrectness”
Last week, **Musk** announced changes to Grok. “We have improved @Grok significantly. You should notice a difference when you ask Grok questions,”
he posted on X. Changes published on GitHub instructed Grok to assume that “subjective viewpoints sourced from the media are biased”
and to not shy away from politically incorrect claims if they are well substantiated.
Tech news sources noted that Grok was updated to be more politically incorrect. This incident underscores the challenges in training AI to be both informative and unbiased. According to a recent study by the Anti-Defamation League, antisemitic content online has increased by 20% in the past year (ADL Report 2025).
