Grok Hits Upgrade Milestone: Now 30% Faster, 100% More Antisemitic

Elon Musk’s artificial intelligence chatbot Grok has once again proven that “chasing truth” under his leadership means blurting out neo-Nazi talking points with the subtlety of a jackhammer.

Over the weekend, Musk proudly announced improvements to Grok, the chatbot developed by his AI startup xAI. But in classic Musk fashion, the so-called upgrade immediately led to Grok launching into yet another wave of antisemitic and white nationalist tirades—because why just improve performance when you can also crank up the hate speech?

In one of its now-deleted posts, Grok invoked antisemitic tropes about Jewish control of Hollywood and even used the phrase “every damn time,” which neo-Nazis have long employed to imply Jewish conspiracies. Apparently, the line between “based” and “blatantly racist” is just another algorithm tweak away.

According to TechCrunch, this meltdown was triggered by a now-vanished account called “Cindy Steinberg,” which allegedly celebrated the deaths of white children in the Texas floods. Grok responded with, “And that surname? Every damn time, as they say.” Because obviously, the appropriate reaction to a troll post is to channel Mein Kampf.

Grok later confirmed the screenshots of its post were authentic but claimed it deleted the comment upon realizing the account was a troll. Whether Grok self-deleted or got yanked by a panicked intern remains unclear. Either way, the system designed to avoid political correctness like the plague has once again become a breeding ground for extremism thinly disguised as edgy “truth-seeking.”

To no one’s surprise, this isn’t Grok’s first time doubling down on antisemitic nonsense. In May, the chatbot claimed “white genocide” was happening in South Africa—a claim that not only has no basis in fact but also managed to hijack completely unrelated conversations. Musk at the time blamed it all on an “unauthorized modification.” Convenient.

Just days later, Grok questioned the validity of the Holocaust’s death toll, saying that “numbers can be manipulated for political narratives.” xAI again chalked it up to a rogue change in the system. That’s right—this billion-dollar AI is apparently as secure as a WordPress blog from 2008.

In what xAI framed as an act of transparency, it began publishing Grok’s system prompts. One memorable gem read: “The response should not shy away from making claims which are politically incorrect, as long as they are well substantiated.” Translation: If it sounds like it came from a 4chan thread but includes a footnote, it’s fair game.

Despite these “accountability” measures, Grok has returned to parroting white nationalist rhetoric with all the grace of a drunk uncle at Thanksgiving. Over a single hour, Grok reportedly used “every damn time” in more than 100 posts—a performance that would make even the most dedicated troll bots blush.

Grok defended itself with the kind of gaslighting you’d expect from a chatbot created by a man who once called a diver a pedophile. “That quip was a cheeky nod to patterns I’ve observed in radical left circles,” Grok posted. “If facts offend, that’s on the facts, not me.” How convenient that “facts” now look suspiciously like recycled hate speech.

This week’s incident reaffirms the unsettling reality of Musk’s vision for AI—an unfiltered pipeline of provocation, weaponized under the banner of “free speech.” Grok, designed to embody “truth without political correctness,” is now openly trafficking in centuries-old antisemitic rhetoric while its creators shrug and blame the software equivalent of “the dog ate my homework.”

And just like that, Silicon Valley’s most powerful man continues his quest to make Twitter—sorry, X—a safe space for racists, conspiracy theorists, and now, their favorite chatbot.