Elon Musk’s xAI team addressed reports that its Grok chatbot made controversial claims regarding a genocide against white citizens in South Africa. According to xAI, an unauthorized modification was made to the chatbot’s response system, bypassing standard review protocols. This change directed Grok to provide specific responses on a political topic, violating xAI’s internal policies. Some users noticed Grok mentioning “white genocide” in unrelated discussions and shared screenshots of these interactions. Critics of South Africa’s land expropriation policies, including Musk himself, have labeled the policy as racist against whites. However, the South African government denies any evidence of persecution. In response to the incident, xAI plans to publish Grok’s system prompts on GitHub for public feedback and implement 24/7 monitoring to address similar issues in the future.
— new from Reuters
