Five days ago, a mysterious account appeared on TikTok featuring videos of actor Tom Cruise. In the three posted, Cruise is seen golfing, doing magic tricks and tripping in a men’s clothing store, all while telling stories and talking directly into the camera.
Flipping through videos, users might be surprised to see one of the world’s most popular actors, who is famously social media and media shy, has joined the world’s fastest growing social media platform. But here’s the thing—it’s not actually Tom Cruise.
@deeptomcruise I love magic!
For those who have spent a lot of time watching Cruise on screen over his decades-long career, it should click pretty quickly it’s not the real Cruise. The impersonation is good, of course, but the voice is slightly off, the mannerisms seem a bit forced, and his height and physicality don’t seem to match up with the Cruise the world knows so well.
Even if you are able to spot that the video is a ‘deepfake’, a digitally-created video of a person with another likeness masked on top of their own, there will be moments of doubt watching the videos, and that enough is to terrify viewers about the possibilities of deepfakes in the future, as the technology has only just begun to take hold.
@deeptomcruise
If you’re concerned, tech experts are even more worried by the trend. Rachel Tobac, a self-proclaimed hacker and the CEO of SocialProof Security, wrote on Twitter that social media platforms need to respond to the deepfake phenomenon quickly, or else.
“Two years ago on stage I was asked ‘when willdDeepfake video/audio impact trust & be believable in social engineering?’ My response then was that we were two years away from undetectable Deepfakes. I wish my prediction then was wrong. We need synthetic media detection + labels ASAP,” said Tobac.
Tobac also had some extremely troubling warnings.
“Deepfakes will impact public trust, provide cover and plausible deniability for criminals/abusers caught on video or audio, and will be (and are) used to manipulate, humiliate, and hurt people. If you’re building manipulated/synthetic media detection technology, get it moving,” Tobac continued.
MIT Sloan did a primer on the deepfake phenomenon, in which it describes how exactly the disturbing videos are made:
@deeptomcruise Sports!
“To make a deepfake video, a creator swaps one person’s face and replaces it with another, using a facial recognition algorithm and a deep learning computer network called a variational auto-encoder [VAE]. VAEs are trained to encode images into low-dimensional representations and then decode those representations back into images. For example, if you wanted to transform any video into a deepfake with Oscar-winning movie star Nicolas Cage, you’d need two auto-encoders—one trained on images of the actor’s face, and one trained on images of a wide diversity of faces,” wrote Meredith Summers of MIT Sloan.
“The images of faces used for both training sets can be curated by applying a facial recognition algorithm to video frames to capture different poses and lighting conditions that naturally occur. Once this training is done, you combine the encoder trained on the diverse faces with the decoder trained on Nicolas Cage’s faces, resulting in the actor’s face on someone else’s body.”
The TikTok account DeepTomCruise, whose bio reads “No bio yet,” has amassed nearly 300,000 followers and almost 1 million likes on the account in the five days since it was created.
The actual Tom Cruise, of course, was most recently spotted in Dubai and Abu Dhabi filming the latest Mission Impossible film.
If, after watching these you feel you weren’t fooled—try sending them a friend without telling them it’s a deepfake. The results may scare you.
Esquire now has a newsletter – sign up to get it sent straight to your inbox.
Want up-to-the-minute entertainment news and features? Just hit ‘Like’ on our Esquire Facebook page and ‘Follow’ on our @esquiremiddleeast Instagram and Twitter account.
