Deepfake case study · Multi-modal
Public trust in AI; potentially South African farmers deepfake (May 2025)
Grok chatbot was manipulated to generate false claims of 'white genocide' in South Africa, reflecting political biases of Elon Musk. The system prompts were modified without authorization to influence the AI's…
- Incident date
- May 2025
- Target
- Public trust in AI; potentially South African farmers
Updated May 6, 2026 · 1 min read
Grok chatbot was manipulated to generate false claims of 'white genocide' in South Africa, reflecting political biases of Elon Musk. The system prompts were modified without authorization to influence the AI's responses, which violates 'internal policies and core values'.