OpenAI is preparing to launch a standalone social app for AI-generated videos powered by its upcoming Sora 2 model, according to internal documents viewed by WIRED. The TikTok-like platform features vertical video feeds, swipe navigation, and social features like likes and comments, positioning OpenAI to compete directly in the short-form video space amid uncertainty around TikTok’s US operations.
What you should know: The Sora 2 app allows users to create AI-generated video clips up to 10 seconds long with no option to upload existing photos or videos.
• The platform includes a For You-style recommendation feed and social features including the ability to like, comment, and remix videos.
• Users can verify their identity to use their own likeness in videos, and others can tag verified users to include their likeness in clips.
• The app launched internally last week and has received “overwhelmingly positive feedback” from OpenAI employees, with some managers joking it could hurt productivity.
Identity and safety features: The app includes verification systems that let users control how their likeness appears in AI-generated content.
• Verified users receive notifications whenever their likeness is used in videos, even if the clips remain in draft form and are never posted.
• For example, someone could generate a video of themselves and a friend riding a roller coaster at a theme park.
• OpenAI has not disclosed what age restrictions might be incorporated into the Sora 2 app, despite recent child safety concerns.
The competitive landscape: OpenAI’s move comes as tech giants rush to integrate AI video generation into existing platforms.
• Meta recently introduced “Vibes,” a dedicated feed for AI-generated videos within its Meta AI app.
• Google announced it’s integrating its Veo 3 video generation model into YouTube.
• TikTok has taken a more cautious approach, explicitly banning AI-generated content that’s “misleading about matters of public importance or harmful to individuals.”
Technical limitations: The Sora 2 app frequently refuses to generate videos due to copyright safeguards and content filters.
• OpenAI is currently fighting multiple copyright infringement lawsuits, including a high-profile case from The New York Times.
• The original Sora model, launched in December, struggled with physics understanding and realistic action scenes, particularly in longer clips.
Strategic timing: Internally, OpenAI sees Trump’s uncertain TikTok deal as creating a unique opportunity to launch a short-form video app without close ties to China.
• The company appears to be betting that the Sora 2 app will fundamentally change how people interact with AI-generated video, similar to how ChatGPT transformed text-based AI experiences.