OpenAI’s Sora App Brings Deepfake Entertainment to Your Phone
OpenAI has just introduced a brand-new app called Sora , and it’s already sparking curiosity, excitement, and a fair bit of concern. The app is designed to let users generate their own digital avatars and scroll through a feed of short, AI-generated videos—much like TikTok, but with a surreal twist: the faces you see, including your own, can be recreated as convincing deepfakes.
Also Read:- Deadly Attack Outside Manchester Synagogue on Yom Kippur
- Fire Alarm Forces Closure of Cobequid Health Centre
The app runs on OpenAI’s latest video generation model, Sora 2 , and it even adds AI-generated sounds to bring the clips to life. At the moment, Sora is available only on iOS and requires an invite code to join, which gives it that exclusive, early-access feel. Upon signing up, users are greeted with a message that sets the tone: “Some videos may depict people you recognize, but the actions and events shown are not real.” In other words, you’re stepping into a playground of AI-created imagination.
Here’s how it works: during the setup, the app asks you to record yourself by speaking a few numbers and turning your head. This process builds a highly accurate digital likeness of you. Once created, that avatar can be dropped into endless scenarios—arguments, adventures, comedy sketches, even animated parodies. You can also decide who gets access to your likeness, whether it’s just you, your friends, or everyone on the app. A neat feature is that whenever your likeness is used, even in someone else’s draft, you can view the video directly from your account.
Early impressions of Sora have shown both its playful potential and its uncanny realism. For example, OpenAI’s CEO Sam Altman’s likeness has been used in popular clips, such as one where he’s caught trying to steal a graphics card. The app’s generation engine creates not only visuals but also dialogue and sound, resulting in short, nine-second clips that can feel unsettlingly real—even when they’re absurd or comedic.
Naturally, this type of technology raises questions about safety and misuse. OpenAI has acknowledged these concerns and implemented several guardrails. The app restricts sexual content, extreme violence, extremist propaganda, and anything that promotes self-harm or disordered eating. Public figures and celebrities are mostly off-limits too, though users have found the system easier to bypass with fictional characters like Pokémon. Still, there are moments where Sora produces errors—voices mismatched to faces, awkward movements—but overall, it’s surprisingly polished.
In many ways, Sora recalls older internet fads like the “Elf Yourself” holiday videos, where people’s faces were slapped onto dancing animations. But unlike those goofy throwbacks, Sora is more dynamic, immersive, and at times eerily convincing. Early testers have described the experience as equal parts entertaining and unsettling. Some users even reported showing Sora-generated clips to friends or partners, who didn’t immediately realize the videos were AI-made.
With Sora, OpenAI isn’t just testing a new app—it’s experimenting with how people might interact with AI entertainment in the future. Whether it becomes the next viral hit or a lightning rod for debate about deepfakes, one thing is certain: Sora blurs the line between imagination and reality in a way that feels both fun and unsettling. And for now, it’s only the beginning.
Read More:
0 Comments