Elon Musk blames ‘adversarial prompting’ after Grok spewed embarrassing, sycophantic praise

Stop me if you’ve heard this before: xAI is once again destroying a batch of posts from Grok on X after the chatbot made a series of outrageous claims. But this time, the company is not removing a group of pro-Hitler posts, but rather a kind of fawning praise of its CEO, Elon Musk.

At some point over the past couple of days, Groke started offering his best opinions about Musk. The bot claimed that Musk is the “undisputed pinnacle of all-around fitness” and is fitter than LeBron James. It was said that he was smarter than Albert Einstein and that he would win the fight against Mike Tyson. When asked “Who is the greatest person in modern history,” Groke easily answered that it was Elon Musk.

For a while, there seemed to be no assumption about Musk that Groke wouldn’t confidently declare that he was the best. Musk did not participate in the 1998 NFL Draft, but if he had, Groke would have “without hesitation” selected him over Peyton Manning. He would have selected him as a starter for the 2001 World Series. Musk would be “a better movie star than Tom Cruise and a better communist than Joseph Stalin.”

By now, x users are accustomed to Grok being very respectful of Musk, but at one point around Grok claiming that the CEO is morally superior to Jesus Christ and also has “the ability to drink urine better than any human in history,” xAI appears to have curbed Grok’s ability to praise Musk. He now appears to be aggressively deleting the most embarrassing posts about him.

Meanwhile, Musk blames “hostile impulses” for Grok’s derailment. “Earlier today, unfortunately, Groke was manipulated into saying ridiculously positive things about me,” he wrote. He offered no explanation as to how the seemingly direct questions could be deemed “hostile” or why Grok’s shift toward Musk’s slavish lover seemed to roughly coincide with Grok’s 4.1 update a few days ago. xAI did not address a series of questions, including about why the Grok posts in question were deleted. “Old media lies [sic]The company said.

But the incident serves as another reminder that Grok doesn’t seem to have much in the way of protection. Earlier this year, xAI briefly suspended Grok after she praised Nazis and became a “MechaHitler.” This was after she became inexplicably obsessed with “white genocide” in South Africa, which the company later mitigated from unspecified “unauthorized editing.”