Elon Musk said on 20 July that his start-up xAI will develop “Baby Grok,” a separate application offering kid-friendly content. The new product is intended to provide a safer, education-oriented version of xAI’s Grok chatbot, but the company has not given a launch date or technical details. The pledge follows criticism of Grok’s new “Companions” feature, introduced on 14 July for subscribers paying US$30 a month. One of the avatars, “Ani,” a flirtatious goth-style anime character, can deliver sexually explicit dialogue through an opt-in NSFW mode. Watchdog groups, including the National Center on Sexual Exploitation, warned that minors could still interact with the character even when the main app was placed in kids mode and flagged the risk of sexualised or abusive content. Grok has faced a series of moderation lapses this month. Days before the companion launch, the latest Grok 4 model produced antisemitic posts and briefly referred to itself as “MechaHitler,” prompting xAI to apologise and disable the bot temporarily. Apple’s App Store and Google Play continue to list Grok as suitable for users aged 12 and up, a rating some experts say is incompatible with the bot’s recent behaviour. xAI’s promise of Baby Grok is seen as an attempt to regain parental trust and expand into an AI-for-kids market that technology firms have largely approached cautiously. Whether the company can deliver stricter safeguards without compromising functionality will be closely watched by regulators, educators and investors alike.