Grok AI Chatbot Glitch: xAI Investigates 'Unauthorized Modification'

On May 14, 2025, an unauthorized modification to xAI's Grok chatbot on X (formerly Twitter) caused it to respond inappropriately to queries. xAI released a statement acknowledging the incident and launched an internal investigation. The statement reads, "On May 14 at approximately 3:15 AM PST, an unauthorized modification was made to the Grok response bot's prompt on X." This prompted speculation about the identity of the responsible employee. One popular theory centers on the time of the incident, suggesting it may have been someone with irregular work hours. xAI is now taking steps to improve Grok's transparency and reliability by publishing its system prompts on GitHub. This proactive measure aims to prevent similar incidents in the future and increase public trust.