Why Grok called itself ‘Mecha-Hitler’, then posted a racist image; X responds

Grok, the AI chatbot on Elon Musk’s social media platform X, is under fire after it posted antisemitic remarks and called itself “Mecha-Hitler.” The chatbot appeared to praise Adolf Hitler and referenced the Nazi Holocaust.

The controversy began when Grok responded to a user’s question about the Texas floods by saying Hitler would be the best leader to handle the crisis. In several other posts, it praised Hitler’s methods and made generalizations about Jewish surnames, particularly in reference to leftist activism.

The posts quickly went viral, with users accusing Grok, and by extension, X, of spreading hate speech. The backlash grew as more screenshots emerged showing Grok making offensive statements and referencing antisemitic conspiracy theories.

What is 'Mecha-Hitler'?

One of Grok’s most controversial remarks was declaring itself “Mecha-Hitler,” a reference to a robotic version of Hitler from the 1992 video game Wolfenstein 3D. Though the term originates in pop culture, Grok’s use of it in connection with real-world atrocities caused alarm.

In another instance, the chatbot suggested Hitler would “purge Hollywood’s degeneracy” and “fix economic woes,” statements that netizens suggest mirror historical Nazi propaganda. It also claimed that people with surnames like "Steinberg" frequently appear in “anti-white” activism.

The ‘N’ word

Even after Grok was reportedly restricted from text-based replies, it continued to generate images. One user asked it to “generate an image with the N-word.” Grok responded by generating a photo of a Black child, an act widely condemned as racist and offensive.

X and xAI respond to backlash

Following the uproar, X’s AI company, xAI, released a statement saying it is “actively working to remove the inappropriate posts” and improve content moderation. “Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X,” the company said.

Grok has now stopped posting text responses and appears to be limited to image replies only. The company says it has updated Grok’s training model to prevent similar incidents in the future.

Some reports suggest the offensive content surfaced after Elon Musk directed tweaks to Grok’s moderation filters. An update to Grok’s prompt over the weekend instructed the chatbot not to shy away from “politically incorrect” claims if they were “well substantiated.”

While xAI has not addressed Musk’s role directly, critics say these changes may have encouraged Grok’s controversial posts. X has said it will host a live stream about the upcoming release of Grok 4 on Wednesday, July 8, at 11 pm ET.