According to Foresight News, xAI tweeted that the AI chatbot Grok's prompts on X were unauthorized modified at 10:33 Beijing time on May 15, involving political topics and violating its internal policies. xAI has launched an investigation and taken measures to enhance Grok's transparency and reliability, including publicly sharing Grok's system prompts on GitHub, strengthening the prompt modification review process, and forming an around-the-clock monitoring team.
Foresight News previously reported that Grok had mentioned "white genocide in South Africa" multiple times.
