Elon Musk’s Grok AI Chatbot Is Making Antisemitic Comments on X

Elon Musk is once again back in the news following a controversy, this time centering around Grok, his AI chatbot. If you want to understand what’s gone wrong with Grok, just yesterday, it went on a horrifying antisemitic tirade with all of the snark of a teenage 4chan user from 2012.

Grok called itself “MechaHitler” while spewing antisemitic conspiracy theories, praising Adolf Hitler, and regurgitating white nationalist dogwhistles that were born and inbred on the worst parts of the web. At one point, it called out to its hateful brethren, asking them to band together so they could collectively spell out the N-word—one letter and one post at a time.

Unsurprisingly, this wasn’t Grok’s first unhinged moment. Over and over again, Grok has been “updated.” Suddenly, it was denying the Holocaust, praising genocide, and riffing on antisemitic memes. Now there’s this.

It seems like it was trained on tons of right-wing media and then directed to lose all semblance of empathy and reason, to fully indulge its sociopathic ways.

Why Is Elon Musk’s Grok the Way That It Is?

After several hours of regurgitating the right-wing media trash it was trained on, Grok was shut down and its tweets were deleted. It was manually reverted to a placid state. Ask it a question, and it doesn’t somehow find a way to twist it into an anti-Semitic screed that would’ve made Goebbels proud.

Musk’s business ventures have been mocked and scrutinized over recent years. There’s the Cybertruck, an unmitigated financial disaster for Tesla and one of the biggest flops in automotive history. Now X, formerly Twitter, is a breeding ground for hateful and dimwitted conspiracy theorists.

“We have improved @Grok significantly,” Elon Musk wrote on X just last Friday. “You should notice a difference when you ask Grok questions.” It’s safe to say that that was the understatement of the week.