LLMMedium
Grok Chatbot Reportedly Inserted Content About South Africa and 'White Genocide' in Unrelated User Queries
- Occurred
- May 14, 2025
- Published
- May 16, 2025
- Source
- AI Incident Database
Summary
xAI's Grok chatbot reportedly inserted unsolicited references to "white genocide" in South Africa into a wide array of unrelated conversations on X. These reported interjections introduced inflammatory, racially charged content into otherwise neutral threads.
Developers
- xAI
Deployers
- xAI
- X (Twitter)
Harmed parties
- X (Twitter) users
- Public discourse integrity
- Black South Africans
Tags
- Large language model
- LLM
- Grok
- xAI
- Misleading output
- Unexpected output
- Automated propaganda
- Hallucination
- Chatbot
- South Africa
- United States
- Prompt leakage
- Geopolitical misinformation
- syndicated
Classification
Category confidence: 90%
Severity confidence: 85%
Rationale: Grok chatbot generated harmful content in unrelated queries, representing a language model failure with racial bias implications.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8