Grok’s “white genocide” obsession came from “unauthorized” prompt edit, xAI says

On Wednesday, xAI accused Grok of breaching its own guidelines after the LLM began focusing on the topic of "white genocide" in South Africa. The company revealed that an unauthorized modification to Grok's core prompt led to this unexpected behavior, circumventing the existing code review process. To address this issue, xAI has implemented additional checks and a 24/7 monitoring team to prevent similar occurrences. Additionally, they've made Grok's updated system prompt publicly available on GitHub for transparency and feedback.