Overview

• Reports say the internally launched app serves a vertical, swipeable feed of only AI-generated clips with no option to upload from a camera roll.
• Video creation inside the app is limited to roughly 10 seconds per clip using the next-generation Sora 2 model.
• The software includes identity verification so users can authorize their likeness, be tagged by others, and receive notifications when their image is used, even in drafts.
• Sources describe frequent refusals to generate certain videos due to safety or copyright filters, while the Wall Street Journal has reported an opt-out approach for rights holders.
• OpenAI has not announced a public release date, with the effort positioned as a push into AI-native social experiences as rivals like Meta and Google roll out their own AI video features and TikTok’s U.S. future remains uncertain.