Sora 2 was officially released by OpenAI as of Oct. 30, 2025. Existing as the successor to OpenAI’s Sora, an app designed for creation of AI videos with simple prompts, Sora 2 boasts the ability to create professional appearing, hyper-realistic videos available at the disposal of everyone with access to the internet.
The main promotion page for Sora 2 displays a multitude of AI generated videos, including a figure skater with a cat balancing on her head, a claymation conductor conducting an orchestra and a cowboy standing on the backs of two horses before falling down in the grass. Some of the videos are humorous, some look professionally animated, all of them look unfathomably realistic.
When observing these videos, I looked for the usual AI indicators — fuzzy display, warped hands and fingers, strange movements. I had come to the jarring realization that I could not distinguish what made these videos fake.
In addition to my unsettling first encounters with Sora 2, I have noticed a bizarre phenomenon of AI-generated “influencers” taking over TikTok. Moreso, some users — such as @notkristen — have observed that these AI influencers are making green screen reaction videos to react to other AI-generated videos, mocking and critiquing them for being fake. They go as far as picking apart the eye movements, the misdrawn limbs and hands in the video in question, all while not actually existing themselves. Fake people are calling out other fake people for being fake.
These AI generated people speak with human-like emotion, show empathetic facial expressions and gesture convincingly. Unless you already know what to look for, you might assume they’re just another creator in your feed. The unsettling truth is that many people have already been fooled.
One of the most unsettling things about Sora 2 is its new cameo feature. It allows people to upload short clips of themselves to create a “self cameo,” allowing their face, voice and mannerisms to appear in any AI-generated video of their choosing. Even more disturbing, it also supports “celebrity cameos,” where historical icons and deceased public figures can be digitally inserted into content. The allotted celebrities currently consist of actors, musicians and athletes like Marilyn Monroe, Micheal Jackson and Muhammed Ali, according to Medium.
Quickly, the line between what is real and what is generated has almost completely vanished. If everyday users can place themselves or their favorite stars into convincing videos, what is stopping this from infiltrating our politics? With Sora 2’s realism, it would be nearly impossible for viewers to know the difference between real and fake politician rhetoric. The same technology that is convincing enough to entertain people on TikTok could be used to spread lies powerful enough to sway an election.
Deepfake videos can make politicians say or do things they never did. Once those clips go viral, they spread faster than the truth ever could. By the time someone fact-checks them, the lie has already done its job. Imagine a fake video of a candidate released days before an election, just enough time to create outrage but not enough to prove it is fake. People will believe what they want to believe, and the rest will just argue online while the truth disappears completely.
It already started with fake robocalls earlier this year, imitating President Biden and telling people not to vote according to NPR. That was only with audio. Now, that same level of deception can exist — but in full video form.
The rise of deepfakes and AI-generated memes is rapidly shifting the terrain of elections around the world. What was once dismissed as fringe satire is now becoming a potent tool of political manipulation. Fake videos, images and audio clips can be produced cheaply and distributed widely, making it harder than ever for voters to trust what they see, according to NPR.
If these AI creations do not win or lose elections on their own, they can erode public confidence, muddy up accountability and give bad actors the ability to plant doubt and confusion at a large scale, according to NPR.
Frankly, younger generations are not ready for how real this technology looks. Because Gen Z grew up with access to the internet unlike their preceding generations, it was once fair to assume Gen Z is too savvy to be fooled by internet scams or fake content. From the Sora 2 videos I have seen, they are nearly indistinguishable from movie scenes, influencer-made Tiktoks and candid clips. If someone who once knew how to distinguish AI from reality is now falling for it, what chance does the average person have?
When people can no longer tell what is real, everything starts to fall apart. It is not just that people believe lies, it is that people stop believing the truth. Real footage can be dismissed as AI. Real victims can be accused of faking evidence. Real journalists can be accused of lacking real sources. Once truth itself becomes optional, accountability disappears. Powerful people will be able to say and do whatever they want because there will always be “proof” that can be manufactured or denied.
This is not just a technological problem, it is an ethical one. Companies releasing these tools need to take responsibility. There should be stricter rules about labeling AI-generated content, especially in politics. Campaigns should not be allowed to use deepfakes under any circumstances, period. Platforms should have systems in place to detect and remove deceptive AI videos before they spread. But it is also on us, as people, part of a community. We have to decide if we still care about truth, about human creativity, about what it means to be authentic, about the irreplaceable value that authenticity brings to our lives.
If we stop caring and allow ourselves to become desensitized to the rise of AI video content, we as Gen Z will be left with a very different world than the one we grew up with. Once deepfakes and AI propaganda become normal, nothing we see or hear will matter anymore. We will live in a world built entirely out of lies, and most people will not even notice when it happens.
Sora AI is not just another move within AI innovation, it is a warning. If we keep treating AI like a harmless experiment, we are going to wake up in a world where reality does not exist. Seeing used to mean believing. Now it means nothing. And if we do not act soon, that is exactly what truth will become — nothing.
