OpenAI is preparing to make another bold move in artificial intelligence, this time aiming to reshape the way people interact with short-form video. The company is working on a TikTok-style social app for AI-generated content, designed to mimic the addictive swipe-and-scroll format that made TikTok a global sensation. But unlike TikTok or Instagram Reels, this platform will not depend on human creators uploading their clips. Instead, every video on the feed will be generated entirely by AI using OpenAI’s powerful new text-to-video model, Sora 2.
The app, which is still undergoing internal testing, is anticipated to offer users an infinite supply of artificial content in the form of vertical films, each lasting roughly ten seconds. OpenAI is developing a platform where the possibilities are almost endless by depending on AI prompts instead of cameras and uploads. This entails viewers browsing through bizarre, funny, or aesthetically beautiful videos that don’t require human intervention to exist. In an AI-first world, it represents another step for OpenAI to establish itself at the forefront of media consumption.
One of the biggest challenges with launching such a platform is managing deepfakes and misuse of likeness. To allay these worries, OpenAI is apparently launching an identity verification system. Users will be able to decide whether or not their likenesses can be used in AI-generated material if they decide to validate their profiles. To ensure transparency before anything is published, they will also be notified if their image is utilized, even in unfinished videos. This strategy aims to reduce the likelihood of impersonation while allowing for imaginative remixing.
Copyright is another thorny issue. Early reports suggest that OpenAI plans to adopt an opt-out framework, which would require rights holders to actively request that their work not be included in AI training or outputs. Given that many creators want for stronger opt-in protections, this approach may prove contentious. However, in an effort to avoid high-profile misuse while continuing to test more general uses of generative video, OpenAI has stated that well-known public people will not be used without express consent.
The timing might work in OpenAI’s favor. Given that TikTok is now dealing with regulatory issues in a number of nations, the firm may view this new endeavor as an opportunity to fill the void and provide a novel, AI-powered substitute. However, the question is whether users will accept a completely fake feed. The sheer inventiveness of machine-generated videos may captivate some, who will like the experience of browsing through an infinite array of AI fantasies. Others could feel alienated because they believe the experience runs the risk of being too manufactured or hollow without actual human creators.
It’s difficult to overlook the potential, though. What it means to “go viral” could be redefined by the TikTok-style social app for AI-generated content, where generative models and algorithms take center stage rather than influencers and content producers. It might develop into a creative playground where concepts come to life at once, making it difficult to distinguish between experimentation and enjoyment.
