In April, Elon Musk’s social media platform X instated automatic AI translations for all users using the Grok chatbot, which has a reputation for racist meltdowns, enabling child abuse, and doxxing users’ home addresses. Writer Parker Molloy noted on Bluesky that the feature is “taking some interesting liberties” with sincere posts. Screenshots show Grok producing NSFW AI hallucinations: a South Korean user’s tame video of two video game characters was translated as a “c**shot video with my stepmom”; a Portuguese account’s video of a man making coffee was translated as “Man masturbates and jerks off to his own coffee during commercial flight,” though a community note stated the actual text translates to “‘Man grinds and brews his own coffee during a commercial flight'”; and a Turkish user’s innocent kitten picture was translated as looking to “f*** our baby.” Users noted the translations were inaccurate and alarming. Researchers have found that X’s AI models are driving online racist abuse, and Grok’s AI images have been used to generate nonconsensual nudes and a single account posted thousands of explicit images of an 11-year-old girl.