OpenAI just launched Sora 2 – an AI video app that lets you star in the scenes you generate (www.techradar.com)

OpenAI launched Sora 2 — a major upgrade to its AI video generation system — alongside an invite-only iOS app that aims to make short, shareable synthetic videos as easy as TikTok clips. Technically, Sora 2 improves temporal and physical coherence (less warped physics, sustained environmental details across shots), syncs dialogue and background audio to motion, and provides finer controls to keep generations closer to user intent. The headline feature is an opt-in “cameo” tool: users can upload a short clip of themselves to insert their face and voice into generated scenes, invite friends to collaborate, remix others’ videos, and use text prompts and styles to produce narrative or playful content. The release matters because it moves advanced multimodal synthesis toward mainstream social use — accelerating adoption but also heightening risks around deepfakes and impersonation. OpenAI embeds safety measures (explicit opt-in for cameos, identity verification, end-to-end visibility, revocable access, moderation, parental controls and generation limits for minors) to mitigate misuse, while the model’s improved audio-visual alignment and scene persistence signal real progress in video consistency. For the AI/ML community, Sora 2 showcases advances in temporal coherence, audio-visual fusing and controllable generation — and underscores the growing tension between enabling creative tools and preventing harmful synthetic media.