Resurrection of legends: how deepfakes are changing cinema
Deepfakes. You may have encountered this phenomenon without even knowing it. Today we will talk about how artificial intelligence is used in filmmaking.
Until recently, digital technologies simply improved special effects, but now they are completely replacing actors. Artificial intelligence is capable of not only erasing wrinkles from a movie star’s face, but also bringing back to the screen those who are no longer alive. Experiments with the «digital resurrection» of stars of the past cause not only admiration, but also questions. How did cinema come to this? It all started with simple special effects, but step by step, technology changed the very approach to filming. Let’s follow this evolution — from the first tricks with miniatures to the era when an actor no longer has to be on the set to play his role.
The film industry has always depended on new technologies, but in the past they only helped actors, not replaced them. In the 20th century, directors used mechanical effects, miniatures and puppet animation. In the 1970s, CGI (Computer-Generated Imagery) technology appeared — computer graphics that allowed audiences to be amazed by realistic spaceships in Star Wars. Then came Motion Capture technologies, where an actor in a special suit transfers movement to a digital character. This is how Gollum, King Kong and Thanos appeared. However, all these technologies required a real person who created the character’s image, spoke and moved for him in the frame. With the advent of deepfakes, everything changed — now you don’t even need to invite an actor to the set.
Deepfakes are the next step in the evolution of cinema, but they also pose a huge challenge to the industry. Now you can recreate any image, and do it faster and cheaper than with animation. Voice, facial expressions, and movements can be completely synthesized, and the actor can be replaced with a computer double. Some people think that deepfakes are just a tool, like makeup or editing. But there is a difference: makeup only emphasizes the personality of the actor, and deepfakes create it anew.
Let’s look at some specific examples. There are studios that are already using fully digital actors instead of real people. This is especially beneficial when it comes to dangerous scenes or when characters need to be rejuvenated. For example, The Irishman used technology to bring back a young De Niro.

Another famous example of a deepfake in cinema is Peter Cushing in Rogue One. The actor died in 1994, but his character, Grand Moff Wilhuff Tarkin, returned to the screen in 2016. To do this, they took another actor, superimposed his movements on a digital model, and added Cushing’s face. It turned out very realistic, but viewers still felt some unnaturalness. This is called the uncanny valley effect, when a character looks almost human, but something is wrong, and this «something» makes us feel uncomfortable or even afraid.

However, the film proved that deepfake actors could play in blockbusters. Remarkably, the studio got permission from Cushing’s heirs, but the actor himself could no longer express his opinion.
Another example is Marlon Brando in Superman Returns. The actor died in 2004, but his character, Jor-El, was recreated using CGI and archival footage. It was a technical breakthrough at the time, but it’s now common practice.
Now you can bring to life and show in a movie any person, even if they have never participated in filming. However, live actors are not always happy with such experiments. Some fear that in the future, studios will stop hiring real people altogether. After all, if you can create an ideal digital character that does not need money, costumes, or even time to rest, why take risks with real actors.
From mechanical tricks to digital immortality, cinema has always kept pace with progress. Deepfakes are the next step, but unlike CGI or motion capture, they challenge the very nature of acting. Whether the future of live-action films will remain or whether we will see an era of virtual stars, time will tell.