On Tuesday, Openai launched Sora 2.0, a new version of its video generation model, with an iOS app that can be used for invitations in the US and Canada. The company highlights advances in physical realism, audio-video synchronization and multi-shot storytelling. The app allows users to create, remix, or view clips generated from text or images using built-in tools to ensure security, traceability, and identity control. Pro versions, web interfaces and developer APIs are planned for the near future.
The launch rekinds debate over intellectual property. According to the Wall Street Journal, Openai warns studios and talent agencies that a protected universe can appear in the creation of Sora unless it explicitly opposes it. The company defends this approach as an extension of the “creative play” that fans have already put into practice, while clearly distinguishing between the use of the work and the use of personal use. Actual people's images and voices can only be generated with prior consent verified by a verified “cameo” system.
Openai presents these safeguards as essential to protecting identity and reputation, and encourages new forms of interaction with characters and the cultural universe. With SORA 2.0, the company wants to establish itself as a key player in AI-generated video creation in a vibrant market by searching for tools that are accessible and secure for both users and rights holders.
