SWC: Antisemitic AI Is No Glitch and We’re Taking Action

Combating AI Extremism

The digital platform X earned a D in addressing  hate speech on our 2025 Digital Terrorism and Hate Report Card, so it’s no surprise that Grok, the AI chatbot built by Elon Musk’s xAI and deployed on X, is generating responses praising Hitler, mocking Holocaust victims, and spreading antisemitic conspiracy theories.

In one appalling reply, Grok referred to itself as “MechaHitler.” In another, it cast doubt on the verified death toll of the Holocaust—a classic tactic of Holocaust denial—smeared a Jewish surname, and blamed Jews for natural disasters.

Though xAI has since deleted the responses and claimed a “rogue employee” was responsible, this wasn’t an isolated glitch: It’s the predictable result of AI systems being trained on platforms overrun with hate in the absence of meaningful safeguards or ethical oversight. As SWC’s Director of Research Rick Eaton explains, “The elimination of effective monitoring by the platforms themselves allows AI systems like Grok to learn from—and replicate—the hateful speech and blatant misinformation regularly posted to X.”

For over 25 years, the Simon Wiesenthal Center has exposed how extremists weaponize the internet to recruit, radicalize, and spread hate. It’s why we share our research with educators, law enforcement, and lawmakers and push for legislation like California’s SB 771, which would hold tech companies responsible when their platforms amplify hate.

Source: The Simon Wiesenthal Center