Hundreds of Mundane AI Apps Are Hiding Sexually Abusive Deepfake Features

WIRED un­co­ve­red some 400 apps and web­si­tes lin­ked to Chi­ne­se-lan­gua­ge de­ve­lo­pers that look like re­gu­lar AI tools but had se­cret fea­tu­res that let users crea­te vio­lent AI-ge­ne­ra­ted ima­ges and video.