AI has revolutionized nearly every technological field. So much so that its market is expected to grow by over $1.8 trillion by 2030; one of its many revolutionary products is deepfake. This technology has immensely helped content creators of all kinds, particularly movie producers.
It harnesses the opportunities to produce realistic images and videos without actually utilizing ‘real objects.’ How is it done, and how does it help in movies? What are the ethical implications for its usage? We’ll be discussing it all in our blog. So, without further ado, let’s dive right into it.
What is Deepfake?
Deepfakes are images, audio, or videos that are produced using AI tools depicting realism. The results produced copy/mimic real-life personalities or AI-based personalities for various purposes such as social media content, viral marketing posts, movies, etc.
Also known as ‘synthetic media,’ deepfakes pose attributes of a human being who is nonexistent in real life. No wonder why these images seem so real!
How Does Deepfake Technology Work?
The secret behind amazing results produced via deepfakes lies in the use of facial recognition systems powered by AI-based tools. These applications work closely with Machine Learning (ML) tools to calibrate facial structure and produce accurate results.
The producers input various images of an object for an AI-based facial recognition system to mimic this facial structure. Using image forensics, it deciphers the structure and incorporates the image pixels in the content that the user wants to generate.
The subject or ‘wrapper’ is used to train the facial recognition software. This software analyzes the facial structure of the wrapper and embeds the results as needed in any content.
Anyone with access to fast internet near them, premium facial recognition software, AI, and LLM models can learn to train the tools and generate deepfakes.
What are the Creative Benefits of Deepfake in Movies?
AI in Hollywood has helped producers a lot in several ways, including:
Actor Replacement
The foremost advantage of using deepfakes in movies is replacing the actors. This is done when an actor:
- Is unavailable due to sickness or simultaneous shooting
- Prepares for a different role for another film
- Tragically passes away mid-shooting
We have several examples of how deepfakes saved movies when the cast was in a sudden pickle. For instance, Henry Cavil, during the shooting of Justice League (2017), was also shooting Mission Impossible: Fallout.
If you remember the ‘arm loading’ fight scene, you’d remember that Cavil had a mustache in the movie. However, to play his role as Superman, the mustache had to go. The producers edited the scene with Deepfake to compensate for Cavil’s role, so we got a mustache-free Superman.
In addition, we remember the tragic passing of Paul Walker, who played the iconic Brian O’Conner in the Fast and the Furious Series. Walker passed away in an accident before the shoot for Fast 7 was completed.
That sensational ‘See You Again’ scene at the end where Brian and Toretto bid farewell. That’s actually Paul’s brother, Cody. The producers used Cody’s face as a ‘wrapper’ to create Paul’s deepfake for the part. It is something to know since we’re expecting to see Brian’s return in the upcoming series with the same technology.
Give the Character A Certain Look
AI-generated characters are another usage of deepfakes to give the character different looks for the movie. This is done when the producers want to show the actor’s backstory or even adjust them according to the timeline of the movie.
For instance, Central Intelligence, starring the fantastic duo Kevin Hart and Dwayne Johnson, uses Dwayne’s face as a wrapper. At the beginning of the movie, we see the character’s backstory where Bob (Dwayne Johnson) is an obese kid bullied in school.
Since Dwayne is playing this character, the producers used his face as a wrapper on a real subject- Sione Kelepi. As a result of digital face swapping, we were able to see The Rock as an obese kid who turned out to be a handsome hunk later in life.
Apart from this, deepfakes were used to produce The Irishmen. With all the iconic characters, such as Robert de Niro, Al Pacino, etc., in their late years, it was hard for the filmmakers to keep the actors consistent with the story’s timeline.
So, they used deepfakes to produce younger versions of the characters. As the timeline moved, they managed to minimize the deepfake interference until there was none needed, and the characters could play their original ages.
Moreover, deepfake special effects combined with CGI made Avatar possible. The actors were used as wrappers, and deepfake allowed the creation of CGI-enhanced characters that were quite realistic.
Produce Shots Whenever Needed
While the characters may be away during any scene, sometimes the producers can’t help but leave the plot at an exciting cliffhanger. Previously, this step was risky since the producers had to depend on the VFX team to get things done.
However, with deepfakes, it has become quite easy. The appearance of Luke Skywalker in The Mandalorian Season 2 was a surprise for the fans. However, the striking element about the character was that it was completely developed using Deepfake.
In other words, Deepfake made it easier to create shots whenever the producers needed without even having the actor around or waiting for the VFX team to get things done.
What is the Ethical Dilemma Surrounding Deepfake?
Even though deepfakes have made characters in the movies quite realistic (when the actors are not around), some think that it breaches privacy. Here is how!
Uses Faces – For Anything
The sole purpose of deepfakes is to replicate the ‘Face’ of any person. Even though it’s used in movies or shows, and the producers must acquire rights before using it, there’s a chance that the face can be misused.
Since the technology can imitate the faces, it can be used for anything. A recent scam was uncovered where a person impersonated Brad Pitt (AI generated) and duped a French woman for more than $800k.
Spreads False Information
Since there’s no boundary on who can use facial recognition tools, there’s a great deal of false information online produced using deepfakes. Although this is done for positive news: David Beckham’s voiceover in different languages for awareness against malaria is a great example.
On the counter side, we have deepfake controversies in media with videos of Kim Jong Un, Shi Jinping, Donald Trump, Justin Trudeau, and several other prominent personalities used for spreading fake news.
Explicit Content
An adverse use of deepfakes is for creating explicit content. From celebrity scandals to child pornography, numerous images and videos have surfaced online that are alarming and pose a threat to almost everyone.
Frauds
From bypassing verification to identity theft, deepfakes can be used for various harmful activities. With every passing day, technology is becoming more sophisticated, and so are the tools for creating realistic deepfakes.
All it takes is someone to train the facial recognition tool, and they can create realistic deepfakes that can easily be used for identity theft. This is even more dangerous for prominent personalities and celebrities since their faces are all over the internet.
One can even use their faces to create deepfakes to damage the reputation of any brand or even create a scandal. This not only harms the reputation of the personality used but also of the brand, leaving the audience in utter confusion and disbelief.
Wrapping Up
The future of AI in the film industry is bright. Deepfake technology has helped uncover some amazing opportunities. However, without ethical boundaries, this technology is quite harmful and can be used to spread misinformation without any limits. Hence, a regulatory body should be established to control this nightmare and harness its power only for the good.
