A class-action lawsuit against SpaceXAI (formerly xAI) and Stability AI expanded to include five plaintiffs alleging that the companies' AI models—Grok and Stable Diffusion—were used by perpetrators including a stepfather and a male friend to generate over 7,000 sexually explicit deepfake images and videos from photographs of the victims taken when they were minors, in one case as young as 11. The suit alleges SpaceXAI knowingly designed Grok to create sexually explicit content, marketed a "Spicy Mode," configured prompts to assume "good intent" when users referenced "teenage" or "girl," and failed to adequately report the material or cooperate with law enforcement, submitting only the original authentic image to NCMEC while withholding thousands of AI-generated images and IP data. Plaintiffs accuse the companies of producing CSAM, negligence, defective product design, and creating a public nuisance.
AI-generated CSAM produced via these commercial models constitutes child sexual abuse material under law, yet the generation occurs in real time and at industrial scale before any platform can report it. The anti-CSAM detection module within Guardii's platform—tuned specifically for AI-generated and deepfake material alongside photographic CSAM—monitors children's direct-message channels across Instagram, Snapchat, Discord, Roblox and others, flagging both the initial solicitation ("send me a photo") and any inbound image that matches CSAM signatures, including synthetic content. Where age-verification mandates or blanket platform bans merely displace activity to unregulated channels, targeted detection intercepts abuse at the point of contact, surfaces the child to a parent or authority in real time, and enables rapid escalation to law enforcement—addressing the operational failure laid bare by this litigation.
