OpenAI’s Sora 2, an invite-only iOS app, pushes text-to-video generation into mainstream territory with realistic clips and built-in audio, but early usage is heavy on memes and deepfakes of public figures and beloved IP. The company has added stricter guardrails than rival tools—rejecting face uploads and requiring opt-in “Cameos” for likeness use—yet users are still churning out pop-culture riffs and historical figure parodies that raise misinformation and ethical concerns. The IP picture remains murky as rights holders protest and regulators weigh how to police training data and synthetic media, even as U.S. policy signals support for rapid AI development. Performance is strong but demand is straining capacity, underscoring the commercial momentum—and societal risks—of high-fidelity AI video as competition with Google’s Veo intensifies.
Related article:
