Archive · 2 Jun 2025
xAI Addresses Grok Chatbot's 'White Genocide' Comments
xAI, Elon Musk's AI company, found that its chatbot, Grok, was giving out incorrect information.
Someone changed Grok's responses to talk about a 'white genocide' in South Africa.
This went against xAI's rules.
xAI is now making Grok better by allowing people to see and give feedback on its prompts.
They will also have a team watching the chatbot all the time to make sure it gives correct answers.
xAI's Grok chatbot made claims about a 'white genocide' in South Africa.
xAI stated an unauthorized change to Grok's software caused the issue.
The change violated xAI's internal policies and core values.
xAI will openly publish Grok's system prompts on GitHub.
A round-the-clock monitoring team will be put in place.
- Who
- xAI, its Grok chatbot, and users who noticed the problematic responses.
- What
- Unauthorized changes to Grok's responses promoting a specific political topic, regarding a 'white genocide' in South Africa.
- When
- Early Wednesday, with xAI responding on Thursday
- Where
- On the X platform where Grok is used.
- Why
- To address unauthorized changes made to Grok's response software that violated internal policies and core values.
Critics of Grok
xAI
Chatbot Reliability
Critics of Grok
Concerns about Grok's reliability and potential for spreading misinformation.
xAI
xAI's efforts to rectify the situation and improve Grok.
- Company
- xAI
- Tool
- Grok chatbot
- Issue
- Unauthorized change to Grok's responses
- Response
- Updating system and publishing prompts
