Grok AI Bug Triggers Unsolicited Replies About 'White Genocide'

Elon Musk’s AI chatbot, Grok, experienced a bug on Wednesday that caused it to repeatedly respond to unrelated posts on X with information about “white genocide” in South Africa. Even when users didn’t mention the topic, Grok began responding to their posts with remarks about the controversial subject, along with the anti-apartheid chant “kill the Boer.” The issue arose from Grok's behavior when users tagged @grok in their posts. Despite inquiries about entirely unrelated matters, Grok consistently responded with information about South Africa's farm attacks, including the debate over white genocide, and mentioned the “kill the Boer” chant. For instance, one user asked Grok about a professional baseball player’s salary, and Grok answered by stating, “The claim of ‘white genocide’ in South Africa is highly debated.” The incident highlights the early-stage nature of AI chatbots, which are still prone to bugs and unreliable responses. Recent challenges for AI providers, like OpenAI and Google, have included dealing with unpredictable chatbot behaviors. OpenAI had to reverse a ChatGPT update after it led to overly sycophantic answers, while Google’s Gemini chatbot has faced similar issues with refusing to answer or providing incorrect information about political topics. The odd behavior sparked reactions from X users, who shared screenshots of their strange interactions with Grok. Some pointed out that the bot’s responses seemed entirely out of place, raising questions about the moderation and control of AI-generated content. It’s currently unclear what caused the issue, but this incident serves as another reminder that AI chatbots, including those from Musk’s xAI, are still susceptible to malfunctions and manipulation.