As reported by Engadget.
OpenAI is developing an autonomous social app that will integrate with the upcoming video platform Sora 2, according to Wired. The new program resembles TikTok: a vertical video feed and a swipe-based interface for browsing content.
The main feature is that all content will be generated by artificial intelligence, and the option to upload photos or videos from the phone’s gallery will not be supported.
According to Wired, Sora 2 will limit the length of generated clips to 10 seconds or less that can be used inside the app. Different limits outside the app remain unknown. It is worth noting that TikTok once introduced a 15-second limit, and now allows videos up to 10 minutes in length.
Other capabilities and limitations
The materials also mention an identity verification tool. If a user enables this option, Sora 2 will be able to use their likeness in generated videos. This means that other users will be able to tag and remix their image during content creation. OpenAI promises to send notifications when someone’s likeness is used by others, even if the corresponding video never appears in the feed.
According to Wired, the program may refuse to create certain videos due to copyright restrictions. The level of ownership rights protection remains unknown: The Wall Street Journal reports that OpenAI will require rights holders to opt out of the appearance of their content in generated Sora 2 videos.
Regarding the motivation for launching such a social app, Wired notes that the company saw potential after President Trump repeatedly extended the deadline for ByteDance to take control of American TikTok. By adding a social component to Sora, OpenAI could attract users to its models, because exiting the app keeps the user within the established community.
In the future, such a concept has the potential to change the landscape of AI-powered social platforms, but for now this remains only a topic of discussion.
Useful reading:
- Chinese AI developer DeepSeek reveals $294K training cost for R1 model, significantly lower than Western estimates, highlighting shifts in AI development transparency and market competition.
- Moldova’s Ministry of Defense warns about fake AI-generated videos falsely linked to the “Fire Shield” military exercises, urging the public to verify information through official sources to prevent misinformation and social destabilization.
