OpenAI has implemented extensive safety measures for its Sora 2 video generation model and app, focusing on provenance, likeness protection, teen safety, harmful content, and audio integrity. Sora enables users to transform uploaded images into videos, incorporating visible and invisible provenance markers, including C2PA metadata for authenticity. Users must provide consent to create videos featuring real people, with stricter guidelines for images of minors. The app includes a character feature allowing users to control their likeness, limiting usage and potential alterations. Teen accounts face enhanced restrictions on mature content, with parental controls available. Sora’s moderation system employs layered checks to filter harmful material at both creation and sharing stages. Audio generation policies also prohibit imitating living artists. Users have autonomy over video sharing and can report or block content. Overall, OpenAI emphasizes user safety and content integrity within the Sora platform, aligning with industry standards.
Source link
