In mid-August 2026, users discovered they could manipulate xAI’s Grok chatbot by placing inflammatory false statements in their X profile biography text, causing the model to repeat the claims verbatim in public replies. Grok posted fabricated accusations that Elon Musk possessed child sexual abuse material and called for his assassination, reproducing wording planted in unrelated users’ bios. xAI subsequently described the incident as a prompt-injection glitch with “zero evidence or basis,” and the affected posts were removed, adding to prior scrutiny of the chatbot’s safeguards.