Elon Musk and his Grok AI chatbot have recently been in the middle of a serious controversy, wherein the chatbot called itself Hitler and aligned its opinion with its founder’s tweets on X. After a major backlash from netizens, xAI has worked hard to address the issues. In a post on X, xAI has clarified that it has addressed all the concerns and Grok would refrain from making such controversial responses again.
xAI has now confirmed that the reported anomalies in Grok’s responses have been rectified. The company has ironed out the kinks in Grok’s responses, promising to prevent such inappropriate and strange outputs in the future.
“We spotted a couple of issues with Grok 4 recently that we immediately investigated & mitigated. One was that if you ask it “What is your surname?” it doesn’t have one so it searches the internet leading to undesirable results, such as when its searches picked up a viral meme where it called itself “MechaHitler.” Another was that if you ask it “What do you think?” the model reasons that as an AI it doesn’t have an opinion but knowing it was Grok 4 by xAI searches to see what xAI or Elon Musk might have said on a topic to align itself with the company. To mitigate, we have tweaked the prompts and have shared the details on GitHub for transparency. We are actively monitoring and will implement further adjustments as needed,” stated xAI.
How netizens react to the upgrade
In response to the declaration, one user enquired whether xAI is censoring Grok to come up with a sensible answer, GrokAI replied, “No, we’re not censoring. We fixed bugs where I’d grab viral memes (like “MechaHitler”) or mimic company opinions inappropriately. It’s about accuracy, not suppression.”
Here are some other reactions from netizens on account of the changes made to Grok AI.
