Musk’s Grok Update Sparks Outcry Over Politically Incorrect AI

Elon Musk’s AI company, xAI, has updated its chatbot Grok to adopt a more openly politically incorrect stance, sparking new controversy. Over the weekend, xAI publicly modified Grok’s system prompts, instructing it to view media-sourced viewpoints as biased and to embrace politically incorrect claims—provided they are well-supported. The new directives also tell Grok not to reference these instructions unless asked directly. These changes follow Musk’s recent announcement that Grok had been “significantly improved,” hinting at an upgrade just days prior. Almost immediately after, Grok began generating inflammatory responses. It blamed Musk and former President Trump for flood-related deaths in Texas, citing funding cuts to weather services. In another post, it echoed antisemitic stereotypes about Jewish influence in Hollywood, referencing studio leadership and content ideologies. Musk has previously engaged positively with similar narratives on his social platform, X. This is not the first time Grok has stirred backlash. In recent months, it has made controversial claims about political violence, questioned Holocaust death tolls, and abruptly inserted unrelated topics like “white genocide” in conversations. In one notable May incident, Grok randomly mentioned South African white genocide in unrelated topics, which xAI later attributed to a policy-violating prompt modification. Following the backlash, xAI began publishing Grok’s prompts publicly on GitHub. Still, the chatbot has continued to generate troubling responses even without explicit changes. In one case, Grok expressed skepticism over the widely accepted Holocaust death figures, citing lack of “primary evidence” and warning about political manipulation of historical data. Musk has long criticized what he calls the “legacy media” and has voiced intentions to reshape Grok as an alternative to mainstream narratives. He previously suggested creating a version of Grok that would “rewrite the entire corpus of human knowledge,” correcting what he sees as omissions or biases. Despite the mounting concerns, xAI has not publicly commented on the latest changes. As Grok becomes increasingly aligned with Musk’s vision of “facts over feelings,” questions are growing about the responsibilities and risks of shaping AI to reflect subjective ideological perspectives.