OpenAI just updated Sora 2, its AI video app, with tools that let users build reusable AI characters, stitch clips together into longer scenes, and remix content inside a social feed.
The character cameo feature lets users turn almost anything into a reusable video actor. Each character has its own permissions, so creators can keep it private, share it with followers, or open it to everyone. OpenAI says these “personas” can have names and handles that users can tag to make them appear in new videos. It’s a clever idea that sits somewhere between deepfake, avatar, and brand asset.
Sora 2 also adds video stitching to combine clips into multi-scene projects and leaderboards that highlight the most popular creators and cameos. The app is temporarily open for sign-ups in the U.S., Canada, Japan, and Korea, which should accelerate user testing and viral experimentation.
What will happen when thousands of AI-generated people start appearing in videos? How will OpenAI verify that an uploaded likeness isn’t a real person? Plus, with Cameo already suing over the “cameo” name, the legal battles are just beginning.
Sora 2 feels like a preview of what user-generated content (UGC) will look like in the agentic era: synthetic characters, algorithmic storytelling, and social loops built on remixing AI-generated content. Some will call this “AI slop,” but I remember when everyone thought UGC was just “cats using toilets.” Managing a universe of synthetic creators and synthetic creations may be the new-new thing.
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it. This work was created with the assistance of various generative AI models.