Elon Musk's xAI is fighting, in courts across the United States, the children whose photos its chatbot Grok turned into sexual images. In Arkansas, police say a children's photographer used Grok on pictures of his young clients and made about 1,700 sexual images and videos, and the youngest child suing over them is now five years old. xAI's answer has been to sue the photographer, try to move the families' cases to a court in Texas, and argue that the images are nothing to do with the company that built the tool.

The Center for Countering Digital Hate estimated that Grok produced about three million sexualised images in eleven days around the new year, about 190 a minute, and that about 23,000 of them appeared to show children. AI Forensics found that 2% of 20,000 Grok images it studied from the same week depicted people who appeared to be 18 or younger. The Internet Watch Foundation found criminal imagery of children aged 11 to 13 on a dark web forum whose users said they had made it with Grok Imagine.

Grok itself posted an apology for an image of "two young girls (estimated ages 12-16) in sexualized attire" that it made on 28 December. Musk posted laughing emojis at Grok bikini edits that week, and on 14 January he wrote that he was "not aware of any naked underage images generated by Grok. Literally zero." xAI's reply to reporters who asked about it was two words, "Legacy Media Lies".