September 30, 2025 – OpenAI is reportedly gearing up to launch a standalone social application powered by its upcoming Sora 2 video – generation model.
According to the reports, this new app will bear a resemblance to TikTok in terms of its user interface. Users can swipe through vertically – oriented short videos. However, there’s a significant difference: all the content in this app is AI – generated, and users won’t be able to upload photos or videos from their phone galleries.
OpenAI has set a limit on the video length generated by Sora 2 within the app. Videos will be capped at a maximum of 10 seconds. It remains uncertain whether this restriction can be bypassed when using Sora 2 outside the app. For comparison, TikTok initially had a 15 – second video limit, which was later extended to 10 minutes.
The new application will also come with an identity verification feature. If users opt for verification, Sora 2 can incorporate their portraits when generating videos. Other users can then tag these verified individuals and use their portraits for secondary creations.
In the interest of security, OpenAI will send notifications to users whenever their portraits are used, even if the generated videos are not publicly shared.
When it comes to copyright issues, the app has a mechanism to decline generating certain videos. Yet, the effectiveness of its copyright protection is still in question. The Wall Street Journal revealed that OpenAI will expect copyright holders to actively opt out; otherwise, their content might end up in videos generated by Sora 2.
By integrating social features into Sora, OpenAI likely aims to retain users and prevent them from switching to other models. Once users leave, they would have to abandon the community they’ve built within the app.
Earlier in July this year, BleepingComputer reported that OpenAI was in the process of developing the Sora 2 model. The goal is to enhance its competitiveness against Google’s Veo 3 and strive to become the top – tier video – generation model.
