Fake Tom Cruise taking over TikTok

The actor has been making a name for himself on TikTok – however, there is one very crucial detail that has people alarmed.

Tom Cruise has had a huge resurgence … on TikTok.

However, the latest videos of the movie star are not actually of him.

The 58-year-old actor is seen in the videos doing simple things like magic tricks, telling jokes and playing golf, but account user @deeptomcruise is actually using “deepfakes” to get Tom Cruise’s face in the video.

RELATED: Tom Cruise’s son now unrecognisable

Deepfakes are manipulated AI technology that can mimic people’s faces in videos.

It is concerning because the technology is very advanced, which essentially means anybody can take another person’s face and make them do whatever they want on camera.

In 2019, Trump followers presented a deepfake of Nancy Pelosi slurring her speech on camera.

On Christmas Day in 2020, British people were also outraged when a deepfake of the Queen dancing on a table emerged.

“That is so creepy,” one person wrote. “You remember that the Queen is a real person, actually alive and all that? Not just a character to play with?”

However, the creepiest thing about this Tom Cruise deepfake is that they are spot on with the actor’s mannerisms.

RELATED: Cruise enlists COVID robots on set

The user has nailed his laugh and stare, but there are a few pointers that give him away: he’s taller than the actor, and some of the lighting of the face up close doesn’t match up.

But despite that, users have voiced their concern at how authentic deepfakes are becoming.

A photographer – Lauren White – commented on one of the videos voicing her worries.

“Deep fakes are getting scary good and taking over TikTok. Every public figure should just be on there with a verified account – even if they don’t want to make content – to make it easier to identify their fakes.”

Deepfakes first popped up in 2017 on Reddit.

The name comes from the conjunction of “deep learning” and “fake”.

It uses facial recognition software and two sets of AI called variational auto-encoders (VAEs).

One is trained on images on a person’s face, while the other has a wide variety of people’s faces in different lighting conditions.

When both are combined, a programmer is able to recreate a person’s face that is able to not only move naturally through the world, but attach itself to another person’s body.

You May Also Like

Leave a Reply

Your email address will not be published. Required fields are marked *