Grok’s ‘white genocide’ modification is a wake-up call for India’s regulators

An unauthorised modification of xAI’s Grok chatbot caused it to start speaking about a genocide that doesn’t exist.

Grok, the chatbot run by Elon Musk’s xAI, took a weird turn last week and began talking about a “white genocide” in South Africa. The company’s response about what caused this to happen raises major concerns about how chatbots can be used to influence narratives, posing a threat to both national security and public safety. 

Here is what happened. Users often reply to posts on social media platform X by tagging the Grok chatbot, asking it for more context about the post and sometimes to even fact-check them. That is a bad idea to begin with, but like many other …