OpenAI will reportedly release a TikTok-like social app alongside Sora 2 (www.engadget.com)

OpenAI is reportedly preparing a standalone social app built around its next-generation video model, Sora 2, that closely mirrors TikTok’s vertical, swipe-to-scroll feed—but with one big twist: all clips in the app will be AI-generated. Wired says the app will only allow machine-created videos (no phone uploads) and will cap in-app Sora 2 clips at 10 seconds; the model’s outside-app limits remain unclear. The product will include an identity-verification option so users can opt in to let Sora 2 generate videos using their likeness, and other users can tag and remix those likenesses. OpenAI will notify people whenever their likeness is used, even for unposted videos. Wired also reports the model will block some copyrighted content, though the Wall Street Journal says rights holders may need to opt out rather than opt in. For the AI/ML community this is significant on multiple fronts: it’s a real-world deployment of large video generative models aimed at mass social consumption, testing content moderation, consent mechanisms, and copyright controls at scale. Technical and policy implications include short-form video generation constraints, how likeness verification and provenance/notification are enforced, the robustness of copyright filtering, dataset and licensing practices, and regulatory scrutiny around deepfakes and synthetic media. The app also represents a strategic move to lock users into an ecosystem where leaving means abandoning community-generated synthetic content, raising questions about platform governance and model stewardship.