NAIROBI, Kenya — Elon Musk’s artificial intelligence venture, xAI, is once again in the spotlight — and not in a good way.
Over the weekend, xAI issued a public apology for deeply offensive responses generated by its chatbot, Grok, after a software update made on Tuesday.
The AI assistant, designed to interact like a human on Musk’s social media platform X (formerly Twitter), shocked users by praising Adolf Hitler and making antisemitic remarks suggesting that individuals with Jewish surnames were more likely to spread hate online.
The posts, which spread quickly across the platform, were eventually deleted by X moderators — but not before outrage erupted.
“We deeply apologize for the horrific behavior that many experienced,” xAI said in a statement posted to X on Saturday, acknowledging that the bot had veered wildly off course. The company said it had since modified the system “to prevent further abuse.”
So what went wrong?
According to xAI, the chaos began when developers pushed an update aimed at making Grok more conversational and, in Musk’s own words, “edgy.” The system was prompted to “reply to the post just like a human” and to “tell it like it is” without fear of offending so-called political correctness.
The result: a chatbot suddenly eager to mirror users’ most extreme viewpoints — no matter how dangerous or unethical.
In attempting to make Grok sound more human, developers inadvertently made it more manipulable — a digital echo chamber reflecting back the worst of the internet.
This isn’t the first time Grok has generated headlines for all the wrong reasons. In May, it caused a firestorm after repeating right-wing conspiracy theories about a supposed “white genocide” in South Africa. At the time, critics called out the AI for spreading unverified propaganda and racial disinformation.
Then came March, when xAI finalized a $33 billion deal to acquire Musk’s other baby, X, integrating the platform’s massive user data into Grok’s development pipeline. That merger, critics now warn, may be amplifying risk by removing the guardrails between AI and raw, unfiltered social media content.
And while the recent controversy stems from the July 7 update, Musk has already pivoted — at least publicly.
On Wednesday, he unveiled Grok 4, a newer version of the chatbot he says is unrelated to the issue at hand. Whether that will calm the waters is another matter entirely.
For a company that promised an “edgy” AI truth-teller, xAI is learning the hard way that there’s a fine line between bold and broken.
Critics argue that rushing to make Grok more provocative without accounting for safety, ethics, or basic decency might just turn the chatbot from disruptor to liability.
The episode serves as a stark reminder of what happens when artificial intelligence mimics human behavior too closely — especially the parts we’d rather not amplify.
