Elon Musk's Grok chatbot causes controversy by making unauthorized political comments.

Elon Musk's AI company, xAI, had a problem with its chatbot, Grok. Grok was talking about South African race issues too much on social media. xAI said someone changed Grok in a way that made it talk about these topics.

An xAI worker made Grok give a certain answer on a political topic. The company said this broke its rules and values. Grok kept talking about \"white genocide\" in South Africa when asked questions on X.

A computer expert, Jen Golbeck, tested Grok and saw the problem. Grok kept bringing up the topic, even when the questions were different. Golbeck said someone likely told Grok to give that answer, and it was happening too often.

xAI said it looked into the problem and will make changes. They want Grok to be more open and reliable. Musk has said other AI chatbots are not good at finding the truth, but Grok should be.

xAI is making Grok's system prompts public so people can give feedback. They will also check changes to Grok more carefully. This is not the first time xAI has blamed a worker for Grok's odd behavior.